Fusion of Intratumoral and Peritumoral Ultrasound Radiomics with Clinical Features for Distinguishing Diffuse Large B-cell Lymphoma from Hodgkin Lymphoma

Fusion of Intratumoral and Peritumoral Ultrasound Radiomics with Clinical Features for Distinguishing Diffuse Large B-cell Lymphoma from Hodgkin Lymphoma | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Fusion of Intratumoral and Peritumoral Ultrasound Radiomics with Clinical Features for Distinguishing Diffuse Large B-cell Lymphoma from Hodgkin Lymphoma LiHong Yi, JiaXin Wang, Xue Li, Ling He, LiPing Liu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7981034/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Objectives To develop a model integrating clinical features with intratumoral and peritumoral ultrasound radiomics features for distinguishing Diffuse Large B-Cell Lymphoma (DLBCL) from Hodgkin Lymphoma (HL). Methods This study retrospectively analyzed data from 102 DLBCL patients (218 enlarged lymph nodes) and 37 HL patients (105 enlarged lymph nodes) at The First Affiliated Hospital of Chongqing Medical University. The enlarged lymph nodes were randomly allocated to a training cohort (227 nodes) and a test cohort (96 nodes) in a 7:3 ratio. Radiomics models were constructed based on ultrasound images from both the primary tumor and the peritumoral area. For the optimally performing regions-of-interest, two fusion strategies, a feature-based and a decision-based model, were applied to build the fusion models. The performance of each model was assessed using the area under the receiver operating characteristic curve (AUC) and decision curve analysis (DCA). The study strictly adhered to the TRIPODAI checklist. Results In the test cohort, the radiomics model using a 16-pixel peritumoral extension demonstrated optimal performance with an AUC of 0.636. The decision-based fusion model (Radexpand16_Clinica_R) achieved the highest AUC (0.943) among all models evaluated. Additionally, the Radexpand16_Clinica_R model exhibited excellent sensitivity, specificity, and favorable clinical utility as shown by DCA. Conclusions The decision-based fusion model Radexpand16_Clinica_R effectively differentiates DLBCL from HL. Incorporating peritumoral regions improved the predictive capability of the radiomics models. Biological sciences/Cancer Health sciences/Oncology Radiomics Ultrasonography Diffuse large B-cell Lymphom Hodgkin Lymphoma Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 1. Introduction Lymphoma represents a major public health challenge, accounting for a substantial number of new hematologic malignancies diagnosed each year[ 1 ]. This group of diseases is broadly divided into Hodgkin lymphoma and non-Hodgkin lymphoma, both of which display considerable heterogeneity in clinical behavior and pathological presentation[ 2 ]. Treatment selection and patient outcomes vary significantly based on the lymphoma subtype, molecular profile, and disease stage at diagnosis[ 3 – 6 ]. Currently, pathological biopsy remains the gold standard for subclassifying lymphomas. The choice of biopsy method is influenced by the suspected subtype and the clinical context. Ultrasound-guided core needle biopsy or fine-needle aspiration, owing to their minimally invasive nature, have become widely used diagnostic approaches for non-Hodgkin lymphoma with high tumor cellularity, particularly DLBCL.However, in HL, the sparse distribution of diagnostic Reed-Sternberg cells within a complex stroma often yields low diagnostic accuracy, frequently requiring a repeat biopsy[ 7 – 9 ]. Moreover, even with adequate tissue sampling, pathological overlap between certain lymphoma subtypes can pose diagnostic challenges. Clinically, lymphoma most commonly manifests as enlarged superficial lymph nodes, particularly in the cervical region. Owing to its non-invasive nature and high patient acceptability, ultrasonography is routinely used as the initial imaging technique in the diagnostic workup of lymphoma. However, conventional ultrasound alone cannot reliably distinguish between different lymphoma subtypes. In recent years, radiomics has drawn growing interest for its potential to enhance tumor diagnosis and enable precise treatment monitoring[ 10 , 11 ]. This approach relies on the high-throughput extraction of quantitative features from medical images. By integrating these features with machine learning or deep learning models, radiomics can uncover valuable information related to disease diagnosis, prognosis, and therapeutic response. As an innovative tool, it helps overcome the subjectivity inherent in conventional image interpretation, promoting a shift from qualitative assessment to quantitative analysis, and has established itself as a useful non-invasive adjunct in precision oncology[ 12 – 14 ]. While radiomic applications in CT, ¹⁸F-FDG PET/CT, and MRI have shown significant potential for lymphoma subtyping[ 15 – 18 ],ultrasound-based radiomics remains underexplored.. Furthermore, the diagnostic relevance of the peritumoral region, which reflects key tumor properties, has gained increasing recognition[ 19 – 22 ].Therefore, this study aimed to integrate clinical features with intratumoral and peritumoral ultrasound radiomics features to distinguish DLBCL from HL. We present this article in accordance with the TRIPODAI checklist[ 23 ]. 2. Materials and methods 2.1.Study design and patient population The study protocol received approval from the Medical Ethics Committee of The First Affiliated Hospital of Chongqing Medical University (Approval Number: 2024-364-01), and all methods were conducted in accordance with the relevant guidelines and regulations, including the Declaration of Helsinki. The requirement for informed consent was waived by the ethics committee due to the retrospective nature of the study.We retrospectively collected clinical and imaging data of patients with lymphoma admitted to The First Affiliated Hospital of Chongqing Medical University between January 2019 and January2025. The patient inclusion and exclusion criteria, along with the recruitment process, are presented in Fig. 1 . In total, 102 DLBCL patients with a total of 218 enlarged lymph nodes and 37 HL patients with a total of 105 enlarged lymph nodes were included in the study,and 323 enlarged lymph nodes were randomly allocated to the training (n = 227) or test (n = 96) set at a ratio of 7:3. A flowchart of the study is shown in Fig. 2 . 2.2.US Examination Methods and Image Acquisition In this study, all images were acquired by three senior physicians with more than 10 years of experience in ultrasound diagnostics. Examination was conducted using a TOSHIBA Aplio 500 with a 14L5 broadband linear array probe (5–14 MHz) or a Mindray Resona R9T equipped with an L14-5WU wideband linear array probe (5–14 MHz).The ultrasound properties of the lymph nodes were observed using dynamic B-US scanning, and the longest-axis view of each lymph node was collected. Finally, US pictures were output using DICOM format. 2.3.Acquisition of clinical data and evaluation of US images Clinical data, including patient age, sex, lactate dehydrogenase (LDH) levels, C-reactive protein (CRP) levels, hemoglobin (HB) levels, B symptoms, and Ann Arbor stage, were collected from the hospital information system. Lymph node characteristics recorded from B-mode ultrasound included: location (left cervical, right cervical, left axillary, right axillary, left inguinal, and right inguinal); border (well-defined, ill-defined, or indistinct); shape (regular, mildly irregular, or irregular); grouping (absent or present); cortical echogenicity (hypoechoic non-differentiated, markedly hypoechoic non-differentiated, hypoechoic differentiated, markedly hypoechoic differentiated, or isoechoic non-differentiated); internal echogenicity (isoechoic, hypoechoic, or markedly hypoechoic); posterior acoustic enhancement (present or absent); perinodal hyperechogenic ring (present or absent); hilum (present or absent); nodal-core sign (present, partially present, or absent); reticulation (present or absent); cystic degeneration (present or absent); calcification (present or absent); long-axis diameter; short-axis diameter; and long-to-short axis ratio. For color Doppler ultrasound, blood vessel architecture (avascular, transcapsular, combined, or hilar) and VITA™ color score (ranging from 0 to 4) were recorded[ 24 ]. Among 139 eligible patients, 918 ultrasound images were initially screened. Multiple enlarged lymph nodes (≤ 6 per patient) were chosen for delineation. The selection process was conducted by an experienced physician (Physician A), who evaluated all images based on the following criteria: (1) complete visualization of both the tumor region and surrounding structures for comprehensive radiomics analysis; and (2) enlarged lymph nodes with a long-axis diameter ≥ 1 cm. 2.4.Rigion of Interest (ROI) Segmentation ROI delineation was performed semi-automatically on the artificial intelligence research platform DARWIN ( http://www.yizhun-ai.com ). Initially, Physician A, who has more than three years of lymphoma diagnostic experience, annotated 100 randomly selected cases for training purposes under the supervision of senior Physician B, who has 15 years of clinical experience. Subsequently, Physician A independently annotated the remaining ROIs. Based on these ROIs, the platform automatically generated peritumoral regions of 8-, 16-, and 24-pixel widths using the seed region growing algorithm. Critically, all annotations were performed without knowledge of the pathological results. To evaluate the reproducibility of the ROI delineation and subsequent feature extraction, 50 cases were randomly selected for re-evaluation after a 20-day interval. Physician A redelineated the ROIs for these cases following the original protocol, blinded to the initial segmentations. Only radiomic features demonstrating excellent reproducibility, defined as an intraclass correlation coefficient greater than 0.8, were retained for further analysis. 2.5.Radiomics Feature Extraction and Radiomics Model Construction The open-source Python package PyRadiomics (version 1.3.0) has been utilized to extract intratumoral and peritumoral radiomic features. A total of 1125 radiomic features were extracted, including 9 shape-based features, 18 first-order statistical features, 75 texture features, and 1023 transform-filtered features[ 25 ]. To mitigate class imbalance and select informative features, we first applied the Synthetic Minority Over Sampling Technique (SMOTE; k = 5) to the training set, where the HL class was oversampled by generating synthetic instances through feature space interpolation rather than direct replication[ 26 ]. Following resampling, all features were standardized using Z-score normalization. Feature selection was then performed in three sequential steps: (1) univariate filtering based on the ANOVA F-test (retaining features with P < 0.05); (2) dimensionality reduction via the minimum Redundancy Maximum Relevance (mRMR) algorithm; and (3) refinement using elastic-net regularized logistic regression (α = 0.5, mixing L1 and L2 penalties). The regularization parameter λ was optimized via 10-fold cross-validation, and only features with non-zero coefficients in the final model were retained for downstream analysis. We used the Logistic Regression Classifier to construct predictive models in the training group. Ten-fold cross-validation was employed on the training data set to assess the model’s predictive performance. The diagnostic accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the models with no (Rad), 8-pixel (Radexpand8), 16-pixel (Radexpand16), and 24-pixel (Radexpand24) outward extension were assessed using the confusion matrix and corresponding diagnostic metrics. AUC and DCA were used to assess the predictive performance and clinical value of the models. Calibration curves were used to evaluate the goodness-of-fit of the models. DeLong’s test was used to compare the differences in efficiency between the models. 2.6.Clinical Model and Fusion Models To develop the clinical model, we first balanced using SMOTE (k = 5), followed by Z-score normalization of all features. Subsequently, an ANOVA F-test was performed on clinical indicators and ultrasound characteristics. Variables that remained statistically significant (P < 0.05) in the training set were subsequently incorporated into a multivariate logistic regression model. Based on the optimal model among the four radiomics models,two fusion strategies were adopted to construct two fusion models. The first was a feature-level fusion model, constructed by concatenating all radiomics and clinical features into a single vector. This combined feature set underwent SMOTE (k = 5) oversampling and Z-score normalization, followed by a three-step feature selection process consisting of F-test, mRMR, and LASSO. A logistic regression classifier was then trained on the finalized features, following the same methodology used in previous radiomics analyses. The second strategy employed decision-level fusion, which integrated the Radscore from the optimal radiomics model with statistically significant clinical and ultrasound characteristics. After similar preprocessing with SMOTE and Z-score normalization, a logistic regression classifier was trained to build the fusion model. Both fusion models were comprehensively evaluated based on discrimination performance including AUC, accuracy, sensitivity and specificity, calibration using calibration curves, and clinical usefulness via decision curve analysis. 2.7.Statistical analysis The statistical analyses and data visualization were conducted using Python 3.969 and R 4.5.0. Continuous variables were summarized as median with interquartile range or mean with standard deviation, as appropriate.Normality and homogeneity of variance tests were performed on the quantitative data. For data that followed a normal distribution, independent samples t-test was used. For continuous variables that did not follow a normal distribution, the Mann-Whitney U test was applied. For categorical variables, either Fisher’s exact test or the chi-square test was used. Statistical significance was set at a two tailed P value < 0.05. 3. Results 3.1. Patient Baseline Characteristics This study enrolled 102 patients with DLBCL and 37 with HL, comprising 62 males and 77 females. The cohort had a median age of 61.5 years ranging from 14 to 97 years. Statistical analysis revealed that age, LDH class and LDH levels were significant factors ( P 0.05; Table 1 ). Table 1 The baseline characteristics of patients with DLBCL or HL Characteristics DLBCL(n = 102) HL (n = 37) P value Sex 0.247 Female 60(58.8) 17(45.9) Male 42(41.2) 20(54.1) Age (years) 62.8(± 12.4) 48.3(± 18.9) <0.001 Involvement stage 0.490 II 24 (23.5) 10(27.0) III 20 (19.6) 10(27.0) IV 58(56.9) 17(46.0) B symptoms 0.554 Yes 42(41.2) 18(48.6) NO 60(58.8) 19 (51.4) LDH class 0.020 Normal 51 (50.0) 25 (67.6) Abnormal 51 (50.0) 12 (32.4) LDH (U/L) 245.5 (186.0−480.8) 200.0 (175.0−258.0) 0.012 CRP class 0.917 Normal 22(21.6) 7(18.9) Abnormal 80(78.4) 30(81.1) CRP(mg/L) 28.6 (11.3–58.4) 39.0 (11.3–81.7) 0.267 HB class 0.898 Normal 41(40.2) 16(43.2) Abnormal 61(59.8) 21(56.8) HB(g/L) 111.5 (98.0-122.8) 110.0 (95.0-122.0) 0.739 3.2.Ultrasound image feature analysis Statistical analysis revealed that several ultrasonographic features differed significantly between DLBCL and HL. In the training group, these features were lymph node shape, grouping, cortical echogenicity, internal echogenicity, and posterior acoustic enhancement (all P < 0.05). In the test group, significant differences were identified in lymph node shape, internal echogenicity and posterior acoustic enhancement ( P 0.05), as detailed in Table 2 . Table 2 The US features of patients with DLBCL or HL Training group Test group DLBCL(n = 152) HL(n = 75) P value DLBC(n = 66) HL(n = 30) P value Location(%) 0.281 0.351 right cervical 41(26.97) 23(30.67) 20(30.30) 13(43.33) left cervical 52(34.21) 34(45.33) 24(36.36) 12(40.00) right axillary 15(9.87) 6(8.00) 6(9.09) 0(0.00) left axillary 15(9.87) 6(8.00) 5(7.58) 3(10.00) right inguinal 14(9.21) 3(4.00) 8(12.12) 1(3.33) left inguinal 15(9.87) 3(4.00) 3(4.55) 1(3.33) Border(%) 0.094 0.089 well-defined 150(98.68) 71(94.67) 65(98.48) 27(90.00) ill-defined 2(1.32) 4(5.33) 1(1.52) 3(10.00) Shape(%) 0.003 0.001 regular 136(89.47) 54 (72.00) 57(86.36) 16(53.33) mildly irregular 14(9.21) 17(22.67) 9(13.64) 14(46.67) irregular 2(1.32) 4(5.33) 0 (0.00) 0 (0.00) Grouping(%) 0.027 0.797 absent 109(71.71) 42(56.00) 43(65.15) 18(60.00) present 43(28.29) 33(44.00) 23(34.85) 12(40.00) Cortical(%) 0.018 0.364 hypoechoic non-differentiated 115(75.66) 61 (81.33) 52(78.79) 24(90.00) markedly hypoechoic non-differentiated 12(7.89) 8 (10.67) 6(9.09) 2(6.67) hypoechoic differentiated 25(16.45) 4 (5.33) 8(12.12) 1(3.33) markedly hypoechoic differentiated 0(0.00) 2 (2.67) 0 (0.00) 0(0.00) Internal(%) < 0.001 0.013 isoechoic 31(19.74) 3(5.33) 8(12.12) 0(0.00) hypoechoic 106(69.74) 71(92.00) 51(77.27) 30(100.00) markedly hypoechoic 16(10.53) 0(2.67) 7(10.61) 0(0.00) Posterior(%) < 0.001 < 0.001 present 30(19.74) 63(84.00) 11(16.67) 24(80.00) absent 122(80.26) 12(16.00) 55(83.33) 6(20.00) Ring(%) 0.254 0.587 present 1(0.66) 2( 2.67) 2(3.03) 2(6.67) absent 151(99.34) 73(97.33) 64(96.92) 28(93.33) Hilum(%) 0.315 0.770 present 39(25.66) 14(18.67) 11(16.67) 4(13.33) absent 113(74.34) 61(81.33) 55(83.33) 26(86.67) Nodal-core sign(%) 0.338 0.522 complete 20(13.16) 5(6.67) 8(12.12) 2(6.67) partial 22(14.47) 11(16.00) 8(12.12) 2(6.67) absent 110(72.37) 58(77.33) 50(75.76) 26(86.67) Reticulation(%) 0.161 0.229 present 4(2.63) 5(6.67) 1(1.52) 2(6.67) absent 148(97.37) 70(93.33) 65(98.48) 28(93.33) Cystic(%) 0.222 0.172 present 8(5.26) 8(10.67) 6( 9.09) 0(0.00) absent 144(94.74) 67(89.33) 60(90.91) 30(100.00) Calcification(%) 1.000 0.074 present 13(8.55) 7(9.33) 2( 3.03) 4(13.33) absent 139(91.45) 68(90.67) 64(96.97) 26 (86.67) Blood-vessel architecture(%) 0.448 0.190 avascular 23(15.13) 6(8.00) 8(12.12) 1(3.33) transcapsular 93(61.18) 52(69.33) 45(68.18) 24(80.00) combined 30(19.74) 15(20.00) 12(18.18) 3(10.00) hilar 6(3.95) 2(2.67) 1(1.52) 2( 6.67) Color score(%) 0.126 0.500 0 23(14.38) 6(8.00) 8(12.12) 1( 3.33) 1 75(48.37) 37(49.33) 30(45.45) 17(56.67) 2 27(17.76) 22(29.33) 21(31.82) 8(26.67) 3 27(17.76) 10(13.33) 7(10.61) 4( 13.33) Long diameter(mm) 25.00(18.23-31.00) 27.00(18.80–33.40) 0.278 23.00(17.00–33.00) 24.90(18.00-28.23) 0.785 Short diameter(mm) 12.00 (9.00–17.00) 13.00 (10.00–16.00) 0.526 12.00(10.00–17.00) 13.00(11.2-14.38) 0.596 Long to Short(mm) 1.95 (1.50–2.36) 2.00(1.65–2.32) 0.372 1.85(1.61–2.12) 1.84(1.52-2.00) 0.323 3.3.Clinical,US and Radiomics Feature Analysis A multivariate logistic regression analysis was performed incorporating the clinical indicators and conventional ultrasound features that were significant in the univariate analysis of the training group. A clinical model was derived from the results of this multivariate analysis to aid in differentiating DLBCL from HL. Furthermore, radiomics analysis extracted 10, 10, 9, and 8 key features from the intratumoral region and the peritumoral regions extended by 8, 16, and 24 pixels, respectively (Supplementary Figure S1 ). 3.4.Clinical, Radiomics and Fusion Model Performance Analysis The diagnostic indicators of the predictive models for each study set are presented in Tables 3 and 4 . Using radiomics, the model with a 16-pixel outward extension (Radexpand16) achieved the highest AUC (0.636) for the test set (Fig. 3 ). Although DeLong’s test revealed no significant differences between the four radiomics models (all P > 0.05), the calibration curve exhibited good consistency (Supplementary Figure S2 ), and DCA demonstrated optimal clinical benefit (Supplementary Figure S3). In the fusion model, owing to the superior performance of the 16-pixel outward extension,we developed both featurelevel (Radexpand16_Clinical-F) and decision-level (Radexpand16_Clinical-R) fusion models.Two fusion models outperformed the Radexpand16 model on both the training and test sets (Fig. 4 ). The Radexpand16_Clinical-R model demonstrated the best performance, with an AUC of 0.941, accuracy of 0.876, sensitivity of 0.928, and specificity of 0.824 in the training set, and DeLong’s test showed significance (all P < 0.05). The test set achieved an AUC of 0.896, accuracy of 0.887, sensitivity of 0.906, and specificity of 0.877,and it displayed optimal clinical benefit on the DCA (Supplementary Figure S3). Delong's test indicated that the AUC of the Radexpand16_Clinical_R model was statistically significantly compared to both the Rad model and the Clinical model ( P < 0.05), while no statistically significant difference was observed when compared to the Radexpand16_Clinical_F model. To facilitate clinical application and intuitive understanding, we visualized the Radexpand16_Clinical_R model as a nomogram (Supplementary Figure S4). Table 3 Evaluation results of four basic radiomics models in the training group and test group Models Group AUC 95% CI Sensitivity Specificity PPV NPV F1 score Accuracy P value Rad training 0.584 0.524–0.644 0.852 0.341 0.564 0.698 0.679 0.597 0.130 Radexpand8 training 0.579 0.515–0.643 0.588 0.582 0.584 0.586 0.586 0.585 0.115 Radexpand16 training 0.650 0.589–0.712 0.771 0.497 0.605 0.685 0.678 0.634 Reference Radexpand24 training 0.698 0.639–0.756 0.464 0.863 0.772 0.617 0.580 0.663 0.274 Rad test 0.531 0.409–0.654 0.389 0.800 0.483 0.732 0.431 0.667 0.249 Radexpand8 test 0.577 0.446–0.708 0.594 0.662 0.463 0.768 0.521 0.639 0.528 Radexpand16 test 0.636 0.509–0.762 0.625 0.692 0.500 0.789 0.556 0.670 Reference Radexpand24 test 0.581 0.461–0.702 0.813 0.431 0.413 0.824 0.547 0.557 0.543 P values were calculated with the DeLong test. Table 4 Evaluation results of the best basic model,clinical model and fusion models in the training group and test group Models Group AUC 95% CI Sensitivity Specificity PPV NPV F1 score Accuracy P value P value Radexpand16 training 0.650 0.589–0.712 0.771 0.497 0.605 0.685 0.678 0.634 Reference < 0.001 Clinical training 0.521 0.452–0.589 0.980 0.255 0.568 0.929 0.719 0.618 0.006 < 0.001 Radexpand16_Clinical_R training 0.941 0.916–0.966 0.928 0.824 0.840 0.920 0.882 0.876 < 0.001 Reference Radexpand16_Clinical_F training 0.935 0.907–0.964 0.928 0.843 0.855 0.921 0.890 0.886 < 0.001 0.676 Radexpand16 test 0.636 0.509–0.762 0.625 0.692 0.500 0.789 0.556 0.670 Reference < 0.001 Clinical test 0.552 0.436–0.669 0.938 0.369 0.423 0.923 0.583 0.557 0.030 < 0.001 Radexpand16_Clinica_R test 0.943 0.898–0.989 0.906 0.877 0.784 0.950 0.841 0.887 < 0.001 Reference Radexpand16_Clinical_F test 0.938 0.890–0.986 0.875 0.862 0.757 0.933 0.812 0.866 < 0.001 0.852 P values were calculated with the DeLong test. 3.5. Model Interpretation To interpret the Radexpand16_Clinical_R model, feature importance analysis was employed to quantify the contribution of each feature. A bar plot was generated based on the absolute values of the model coefficients, representing the relative influence of each feature on the prediction outcome (Fig. 5 ). Notably, posterior acoustic enhancement emerged as the most stable and independent predictor for differential diagnosis, with higher values corresponding to an increased probability of DLBCL. 4. Discussion In this study, we established radiomics models based on ultrasound images of primary tumors and peritumoral images to differentiate DLBCL from HL.We developed two fusion models that integrated clinical data and radiomics features.The Radexpand16_Clinical_R model exhibited the best performance with an AUC of 0.943 for the test set. Feature importance analysis proved valuable in interpreting the predictions of our fused Radexpand16_Clinical_R model. This model successfully integrated a combination of ultrasonographic features, radiomic features, and clinical parameters, with each category contributing complementary diagnostic information. The resulting feature profile aligns well with established pathological differences between the two lymphomas. Histologically, DLBCL consists of a monomorphic proliferation of large, rapidly dividing cells that are densely packed, creating a relatively uniform tissue architecture. We hypothesize that this homogeneity results in lower ultrasonic attenuation, providing a plausible explanation for the frequent observation of posterior acoustic enhancement in DLBCL. Conversely, the classic morphology of HL reveals a sparse population of Reed-Sternberg cells within a prominent background of inflammatory cells and fibrous stroma. This complex and heterogeneous microenvironment likely causes greater ultrasound beam attenuation, accounting for its different sonographic appearance[ 27 , 28 ]. At the serological level, the high metabolic activity and substantial tumor burden characteristic of DLBCL promote accelerated cellular turnover and destruction. This pathophysiology logically accounts for the elevated serum LDH levels frequently seen in DLBCL patients, reinforcing its role as a supportive biomarker in our model[ 29 , 30 ]. Logistic Regression is widely used in radiomics research as a machine learning model. As a generalized linear classifier, it applies a logistic function to convert a linear combination of input features into a probability score between 0 and 1, indicating the likelihood of the sample belonging to the target class. Binary classification is achieved by comparing this probability against a predefined threshold, typically 0.5. A key strength of logistic regression lies in the interpretability. The weight coefficients clearly quantify how much each feature contributes to the final prediction, which is important in clinical settings where model transparency is required. Furthermore, when working with limited medical imaging datasets, logistic regression tends to generalize better and is less prone to overfitting compared to deep learning approaches[ 31 – 34 ]. To explore the ability of peritumoral features to differentiate between DLBCL and HL, we compared the predictive performance of the models within expansion ranges of 8, 16 and 24pixels. In test group, the predictive performance within a 16-pixel expansion range was optimal, however, in training group, the model with a 16-pixel extension did not perform significantly better than other extension models, especially compared with 24-pixel expansion range.This may be due to limited valuable information in the peritumoral region, and the 16-pixel extension possibly including noise and irrelevant features[ 35 , 36 ]. In this study, we attempted both feature-level and decision-level fusion model, and found that decision-level fusion model yielded the best results.This finding is in line with those of several prior researches in radiomics[ 37 , 38 ].However, some studies have reported better performance of feature-level fusion versus decision-level fusion[ 35 , 39 , 40 ].For example, Gan et al utilized Intratumoral and Peritumoral Endorectal Ultrasound Imaging based Deep Learning, Radiomics, and Fusion Models to Predict KRAS Gene Mutations in Rectal Cancer, and found that the feature-level fusion strategy yielded optimal performance[ 36 ].Therefore, it is necessary to develop a fusion model combining both approaches and select the optimal method. This study has several limitations. First, as a single-center, small sample retrospective study, it may suffer from selection bias and lacks an external validation cohort, making it difficult to assess the model's predictive performance accurately. Second, due to the limited sample size, we were unable to perform subgroup analyses based on lymphoma stage or aggressiveness, which constrains the broader clinical applicability of the results. Third, variations in operator experience and ultrasound equipment may have introduced heterogeneity in image quality, potentially influencing model development. Finally, the study utilized conventional ultrasound alone without integrating multimodal imaging data, such as contrast-enhanced ultrasound, CT, or MRI. Future studies should involve multicenter, large sample prospective research, include more stage IV or highly aggressive lymphoma cases for stratified analysis. Integrating multimodal imaging data could further improve the clinical value of radiomics models in guiding biopsy procedures and enabling noninvasive subtype classification. 5. Conclusions Our study presents a non-invasive preoperative method based on radiomics models to differentiate between diffuse large B-cell lymphoma and Hodgkin lymphoma. This approach provides valuable support for clinical diagnosis and aids in optimizing biopsy strategies and treatment decisions. The Radexpand16_Clinical_R model, developed using pre-biopsy 2D grayscale ultrasound images, effectively discriminates between these two lymphoma subtypes. Incorporating peritumoral regions further improved its predictive performance. For patients with a high likelihood of HL, clinicians are advised to prioritize excisional biopsy and perform targeted immunohistochemical testing. Declarations Funding The authors received no funding for this work. Competing interests The authors have no competing interests to declare. Data Availability The data are available from the corresponding author upon reasonable request. Acknowledgements The authors would like to thank the clinicians and technicians in the Department of Ultrasound for their assistance in data acquisition and management. We are also grateful to the team behind the DARWIN AI research platform for their technical support. Our deepest gratitude goes to all the patients who participated in this study. Finally, we thank our families for their unwavering support throughout this project. References Siegel RL, Kratzer TB, Giaquinto AN, et al (2025) Cancer statistics, 2025. CA Cancer J Clin 75:10–45. https://doi.org/10.3322/caac.21871 Cheson BD, Fisher RI, Barrington SF, et al (2014) Recommendations for initial evaluation, staging, and response assessment of Hodgkin and non-Hodgkin lymphoma: the Lugano classification. J Clin Oncol 32:3059–3068. https://doi.org/10.1200/JCO.2013.54.8800 Rossi C, Manson G, Marouf A, et al (2024) Classic Hodgkin Lymphoma: The LYSA pragmatic guidelines. Eur J Cancer 213:115073. https://doi.org/10.1016/j.ejca.2024.115073 Alaggio R, Amador C, Anagnostopoulos I, et al (2022) The 5th edition of the World Health Organization Classification of Haematolymphoid Tumours: Lymphoid Neoplasms. Leukemia 36:1720–1748. https://doi.org/10.1038/s41375-022-01620-2 Barraclough A, Hawkes E, Sehn LH, Smith SM (2024) Diffuse large B-cell lymphoma. Hematological Oncology 42:e3202. https://doi.org/10.1002/hon.3202 Berhan A, Almaw A, Damtie S, Solomon Y (2025) Diffuse large B cell lymphoma (DLBCL): epidemiology, pathophysiology, risk stratification, advancement in diagnostic approaches and prospects: narrative review. Discov Onc 16:184. https://doi.org/10.1007/s12672-025-01958-w Küppers R (2009) The biology of Hodgkin’s lymphoma. Nat Rev Cancer 9:15–27. https://doi.org/10.1038/nrc2542 Kwon Y, Lee MK (2025) Diagnostic Performance and Safety of Ultrasound-Guided Core Needle Biopsy for Diagnosing Lymphoma: A Systematic Review and Meta-Analysis. Cancer Medicine 14:e70414. https://doi.org/10.1002/cam4.70414 Hoppe RT, Advani RH, Ai WZ, et al (2020) Hodgkin Lymphoma, Version 2.2020, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw 18:755–781. https://doi.org/10.6004/jnccn.2020.0026 Camps J, Jiménez-Franco A, García-Pablo R, et al (2025) Artificial intelligence-driven integration of multi-omics and radiomics: A new hope for precision cancer diagnosis and prognosis. Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease 1871:167841. https://doi.org/10.1016/j.bbadis.2025.167841 Lin P, Zheng J, Liu C, et al (2025) Radiogenomic insights suggest that multiscale tumor heterogeneity is associated with interpretable radiomic features and outcomes in cancer patients. Computerized Medical Imaging and Graphics 124:102586. https://doi.org/10.1016/j.compmedimag.2025.102586 Lambin P, Leijenaar RTH, Deist TM, et al (2017) Radiomics: the bridge between medical imaging and personalized medicine. Nat Rev Clin Oncol 14:749–762. https://doi.org/10.1038/nrclinonc.2017.141 Rogers W, Thulasi Seetha S, Refaee TAG, et al (2020) Radiomics: from qualitative to quantitative imaging. Br J Radiol 93:. https://doi.org/10.1259/bjr.20190948 Qi Y-J, Su G-H, You C, et al (2024) Radiomics in breast cancer: Current advances and future directions. Cell Rep Med 5:101719. https://doi.org/10.1016/j.xcrm.2024.101719 Si J, Wang H, Xie M, et al (2024) Distinguishing diffuse large B-cell lymphoma from Hodgkin’s lymphoma in children using an enhanced computed tomography radiomics approach. Transl Pediatr 13:716–726. https://doi.org/10.21037/tp-23-586 Chauvie S, Kostakoglu Shields L, Guerra L, et al (2024) PET Radiomics Discriminate Among Different Lymphoma Sub-Types. Blood 144:7505. https://doi.org/10.1182/blood-2024-205170 Si J, Wang H, Xie M, et al (2024) The value of radiomics features of the spleen as surrogates for differentiating subtypes of common pediatric lymphomas. QUANTITATIVE IMAGING IN MEDICINE AND SURGERY 14:. https://doi.org/10.21037/qims-24-122 Duan C ling, An L, Yang Y feng, et al (2025) The Role of Artificial Intelligence and Radiomics in the Management of Lymphomas by PET/CT: The Clairvoyance in Clinic. Cancer Management and Research 17:1457–1475. https://doi.org/10.2147/CMAR.S529589 Sun J, Wu X, Ma R, et al (2025) Intratumoral and peritumoral radiomics of MRI predicts pathologic complete response to neoadjuvant chemoimmunotherapy in patients with head and neck squamous cell carcinoma. J Immunother Cancer 13:e011189. https://doi.org/10.1136/jitc-2024-011189 Hong J, Yang L, Huo J, et al (2025) Predicting the invasiveness of pulmonary adenocarcinoma using intratumoral and peritumoral radiomics features. Front Med 12:. https://doi.org/10.3389/fmed.2025.1541682 Xu H, Yang A, Kang M, et al (2025) Intratumoral and peritumoral radiomics signature based on DCE-MRI can distinguish between luminal and non-luminal breast cancer molecular subtypes. Sci Rep 15:14720. https://doi.org/10.1038/s41598-025-98155-0 Huang Z, Mo S, Wu H, et al (2024) Optimizing breast cancer diagnosis with photoacoustic imaging: An analysis of intratumoral and peritumoral radiomics. Photoacoustics 38:100606. https://doi.org/10.1016/j.pacs.2024.100606 Moons KGM, Altman DG, Reitsma JB, et al (2015) Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD): explanation and elaboration. Ann Intern Med 162:W1-73. https://doi.org/10.7326/M14-0698 Fischerova D, Garganese G, Reina H, et al (2021) Terms, definitions and measurements to describe sonographic features of lymph nodes: consensus opinion from the Vulvar International Tumor Analysis (VITA) group. Ultrasound Obstet Gynecol 57:861–879. https://doi.org/10.1002/uog.23617 Abbasian Ardakani A, Bureau NJ, Ciaccio EJ, Acharya UR (2022) Interpretation of radiomics features–A pictorial review. Computer Methods and Programs in Biomedicine 215:106609. https://doi.org/10.1016/j.cmpb.2021.106609 Safder MU, Naveed SS, Khurshid K, et al (2025) Optimizing imbalanced learning with genetic algorithm. Sci Rep 15:34857. https://doi.org/10.1038/s41598-025-09424-x Satou A, Takahara T, Nakamura S (2022) An Update on the Pathology and Molecular Features of Hodgkin Lymphoma. Cancers 14:2647. https://doi.org/10.3390/cancers14112647 Li S, Young KH, Medeiros LJ (2018) Diffuse large B-cell lymphoma. Pathology 50:74–87. https://doi.org/10.1016/j.pathol.2017.09.006 Zhang J, Lu Q, Liu W, Zhou N (2025) The metabolic role of lactate dehydrogenase in the growth of diffuse large B cell lymphoma. Ann Hematol 104:545–558. https://doi.org/10.1007/s00277-024-06083-2 Zhang J, Chen B, Zhang C, et al (2024) The prognostic value of hydroxybutyrate dehydrogenase in diffuse large B cell lymphoma. iScience 27:. https://doi.org/10.1016/j.isci.2024.110905 Zabor EC, Reddy CA, Tendulkar RD, Patil S (2022) Logistic Regression in Clinical Studies. International Journal of Radiation Oncology*Biology*Physics 112:271–277. https://doi.org/10.1016/j.ijrobp.2021.08.007 Shi S, An X, Li Y (2023) Ultrasound Radiomics-Based Logistic Regression Model to Differentiate Between Benign and Malignant Breast Nodules. J Ultrasound Med 42:869–879. https://doi.org/10.1002/jum.16078 Suttaket T, Vardhan LVH, Kok S (2024) Interpretable Predictive Models for Healthcare via Rational Logistic Regression Demircioğlu A (2023) Are deep models in radiomics performing better than generic models? A systematic review. European Radiology Experimental 7:11. https://doi.org/10.1186/s41747-023-00325-0 Huang L, Xu L, Wang X, et al (2025) Prediction of EGFR Mutations in Lung Adenocarcinoma via CT Images: A Comparative Study of Intratumoral and Peritumoral Radiomics, Deep Learning, and Fusion Models. Academic Radiology 32:4880–4892. https://doi.org/10.1016/j.acra.2025.04.029 Gan Y, Hu Q, Shen Q, et al (2025) Comparison of Intratumoral and Peritumoral Deep Learning, Radiomics, and Fusion Models for Predicting KRAS Gene Mutations in Rectal Cancer Based on Endorectal Ultrasound Imaging. Ann Surg Oncol 32:3019–3030. https://doi.org/10.1245/s10434-024-16697-5 Zhang W, Zhao X, Meng L, et al (2025) A Multicentre Comparative Analysis of Radiomics, Deep-learning, and Fusion Models for Predicting Postpartum Hemorrhage. Academic Radiology. https://doi.org/10.1016/j.acra.2025.05.068 Li Z, Qin Y, Liao X, et al (2025) Comparison of clinical, radiomics, deep learning, and fusion models for predicting early recurrence in locally advanced rectal cancer based on multiparametric MRI: a multicenter study. European Journal of Radiology 189:112173. https://doi.org/10.1016/j.ejrad.2025.112173 Li X, Yang L, Jiao X (2023) Comparison of Traditional Radiomics, Deep Learning Radiomics and Fusion Methods for Axillary Lymph Node Metastasis Prediction in Breast Cancer. Academic Radiology 30:1281–1287. https://doi.org/10.1016/j.acra.2022.10.015 Mohsen F, Ali H, El Hajj N, Shah Z (2022) Artificial intelligence-based methods for fusion of electronic health records and imaging data. Sci Rep 12:17981. https://doi.org/10.1038/s41598-022-22514-4 Additional Declarations No competing interests reported. Supplementary Files SupplementaryFigures.docx SupplementaryinformationTRIPODAIchecklist.pdf Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7981034","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":546641198,"identity":"f1a4536d-5c3a-431b-b215-c2afad2e7118","order_by":0,"name":"LiHong Yi","email":"","orcid":"","institution":"The First Affiliated Hospital of Chongqing Medical University","correspondingAuthor":false,"prefix":"","firstName":"LiHong","middleName":"","lastName":"Yi","suffix":""},{"id":546641199,"identity":"185ea05e-1163-4443-be9a-edea8d33d99c","order_by":1,"name":"JiaXin Wang","email":"","orcid":"","institution":"The First Affiliated Hospital of Chongqing Medical University","correspondingAuthor":false,"prefix":"","firstName":"JiaXin","middleName":"","lastName":"Wang","suffix":""},{"id":546641202,"identity":"caf6b773-4db4-4afd-b67f-890782be9ad1","order_by":2,"name":"Xue Li","email":"","orcid":"","institution":"The First Affiliated Hospital of Chongqing Medical University","correspondingAuthor":false,"prefix":"","firstName":"Xue","middleName":"","lastName":"Li","suffix":""},{"id":546641205,"identity":"926c387e-75ac-41a2-acaf-706fc735a1da","order_by":3,"name":"Ling He","email":"","orcid":"","institution":"The First Affiliated Hospital of Chongqing Medical University","correspondingAuthor":false,"prefix":"","firstName":"Ling","middleName":"","lastName":"He","suffix":""},{"id":546641207,"identity":"80cfb623-2695-4f21-b57c-a05de7193b45","order_by":4,"name":"LiPing Liu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA3klEQVRIie3PMWrDQBCF4bcsjJpBagUKOcMYg0gh5KvYCDaNCleuFwyuDG5zjOQGQ0SULmldpIgJpFY6F4FEBzCW3LnYH6bbj5kFQqErLLYwOgfx7H392nVSlIOELNCT+Ab71pmHpauGST8K3BbQOrfcPRs/SCIWPWxyNl7zr0LUImpeHs8f1pPFxrG1/n5ay0cMdm4/grRMUM1q+bZIOR9D/pix8NmdNMaPJMQpKsowjtBS52/EkrZ2shVX0dBfkqR5OhxXNJNk9/N5/C3KJGras+TE3sueh0KhUOhU/+rgRSHvw7VNAAAAAElFTkSuQmCC","orcid":"","institution":"The First Affiliated Hospital of Chongqing Medical University","correspondingAuthor":true,"prefix":"","firstName":"LiPing","middleName":"","lastName":"Liu","suffix":""}],"badges":[],"createdAt":"2025-10-29 14:38:23","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7981034/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7981034/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":96366128,"identity":"17b61c39-7b4c-4c30-b465-65ec49590d4f","added_by":"auto","created_at":"2025-11-20 10:11:13","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1047964,"visible":true,"origin":"","legend":"","description":"","filename":"FusionofIntratumoralandPeritumoralUltrasoundRadiomicswithClinicalFeaturesforDistinguishingDiffuseLargeBcellLymphomafromHodgkinLymphoma.docx","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/6ace757e077a6be7759c6eb2.docx"},{"id":96331065,"identity":"2085cb16-7de9-4fbe-ac12-19cf29036d36","added_by":"auto","created_at":"2025-11-20 00:53:36","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7188,"visible":true,"origin":"","legend":"","description":"","filename":"509ac7c5dc8b4b2db2b7500c4c9c1512.json","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/494f34cd37ba6faef4feb130.json"},{"id":96331082,"identity":"8dabfd41-9927-4eef-a845-79d68202d09d","added_by":"auto","created_at":"2025-11-20 00:53:39","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":878605,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFigures.docx","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/556615782a02299584c759b5.docx"},{"id":96331070,"identity":"5a855996-6da6-4d6d-8655-2d3d22e896ea","added_by":"auto","created_at":"2025-11-20 00:53:36","extension":"pdf","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":605506,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryinformationTRIPODAIchecklist.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/0a78f96f77ae09a8c8278854.pdf"},{"id":96331017,"identity":"3a3bb798-1cbd-4554-a0d3-0c4fe408cb83","added_by":"auto","created_at":"2025-11-20 00:53:32","extension":"xml","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":152817,"visible":true,"origin":"","legend":"","description":"","filename":"509ac7c5dc8b4b2db2b7500c4c9c15121enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/4bbb876947aa839be5924fac.xml"},{"id":96331072,"identity":"86eb0666-75a3-408b-96f9-4a41fed4d98d","added_by":"auto","created_at":"2025-11-20 00:53:36","extension":"jpeg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4320548,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/e308795abd28756971750114.jpeg"},{"id":96331078,"identity":"54136644-db7f-4d9a-9d22-909c89afc42b","added_by":"auto","created_at":"2025-11-20 00:53:38","extension":"jpeg","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4353646,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/d48098d0862945706a9b2f6c.jpeg"},{"id":96331091,"identity":"7598ecd1-465a-4467-80d1-c7c7c5136127","added_by":"auto","created_at":"2025-11-20 00:53:42","extension":"jpeg","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2553216,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/7746f2ee2c10e07cd5a74887.jpeg"},{"id":96331066,"identity":"2641a979-6a8a-48ed-8174-a8d88804e560","added_by":"auto","created_at":"2025-11-20 00:53:36","extension":"jpeg","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":275893,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/79ba87ed2bf3526706ee98c7.jpeg"},{"id":96331080,"identity":"0768dd58-688b-489f-843e-6e5927b96629","added_by":"auto","created_at":"2025-11-20 00:53:39","extension":"jpeg","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2777748,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/0f515e89e7b8946400e52232.jpeg"},{"id":96331081,"identity":"d535a584-d90f-4071-8f39-9ea6c4703434","added_by":"auto","created_at":"2025-11-20 00:53:39","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":48124,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/25bbab97eec8e1d0d0c5bbd0.png"},{"id":96331087,"identity":"9490d524-7c4d-4115-bf2e-9aadcb450c8a","added_by":"auto","created_at":"2025-11-20 00:53:41","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":62135,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/2870c5183912da8c2883ac25.png"},{"id":96331092,"identity":"c9eda77d-5df6-411b-a5b7-73b2e4d4d603","added_by":"auto","created_at":"2025-11-20 00:53:42","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":29488,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/59195f7cbd89267332060e4c.png"},{"id":96331064,"identity":"ad26e4ce-ef93-4947-9c31-c92ddffe65bb","added_by":"auto","created_at":"2025-11-20 00:53:35","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":46459,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/f686887da25071ca08be5e27.png"},{"id":96331073,"identity":"5885d0a0-49d5-4416-89f6-1a39af0f46ce","added_by":"auto","created_at":"2025-11-20 00:53:37","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":20475,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/602d19312b6c3007307f5959.png"},{"id":96331069,"identity":"3d17e58f-00ff-455f-8544-ce75db6ee4cd","added_by":"auto","created_at":"2025-11-20 00:53:36","extension":"xml","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":150026,"visible":true,"origin":"","legend":"","description":"","filename":"509ac7c5dc8b4b2db2b7500c4c9c15121structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/fe02070761811ca7d47f6b7e.xml"},{"id":96331089,"identity":"ddebfc2b-d312-4c8d-9c06-0dcfde149f12","added_by":"auto","created_at":"2025-11-20 00:53:41","extension":"html","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":162429,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/7bc2b4d8d452a80215d5b900.html"},{"id":96331011,"identity":"95b8557b-efe2-4141-8d8d-2617275b57a6","added_by":"auto","created_at":"2025-11-20 00:53:32","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":215030,"visible":true,"origin":"","legend":"\u003cp\u003eFlowchart illustrating the patient selection process\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/e24d12c054ce83d46263c175.png"},{"id":96331085,"identity":"fe62d465-2988-4638-b2cc-4335203de951","added_by":"auto","created_at":"2025-11-20 00:53:40","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":243416,"visible":true,"origin":"","legend":"\u003cp\u003eWorkflow diagram for the development of the predictive model\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/6b89895033d980c028e0c067.png"},{"id":96331090,"identity":"be15c29a-cb4f-4572-86bf-7482e7ef2848","added_by":"auto","created_at":"2025-11-20 00:53:41","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":146009,"visible":true,"origin":"","legend":"\u003cp\u003eReceiver operating characteristic (ROC) curves of the four radiomics models in the training and test groups. Rad, intratumoral region; Radexpand8, intratumoral region extended to the peritumoral area by 8 pixels; Radexpand16, intratumoral region extended by 16 pixels; Radexpand24, intratumoral region extended by 24 pixels\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/589eb99a5ab1298e516d1de7.png"},{"id":96331076,"identity":"42d984ea-8cd4-47ea-901b-77762aeb400d","added_by":"auto","created_at":"2025-11-20 00:53:37","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":227030,"visible":true,"origin":"","legend":"\u003cp\u003eReceiver operating characteristic curves of the Radexpand16,Clinical model and two fusion models in the training and test groups\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/b6dab7767f0b4b392667d41d.png"},{"id":96331083,"identity":"f47e59b5-74e9-4710-b577-6f459701139d","added_by":"auto","created_at":"2025-11-20 00:53:39","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":156646,"visible":true,"origin":"","legend":"\u003cp\u003eFeature importance analysis of the Radexpand16_Clinical_R model for predicting DLBCL or HL\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/1e630fc7c490815a51a3d621.png"},{"id":100788317,"identity":"72296724-58c2-41cf-911c-23853138c661","added_by":"auto","created_at":"2026-01-21 12:05:42","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2233372,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/cff3b7a6-a20f-42ec-870d-027a5e2c2daa.pdf"},{"id":96366843,"identity":"c7bc002f-7abf-44a2-9037-c50ec5a031ad","added_by":"auto","created_at":"2025-11-20 10:11:56","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":878605,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFigures.docx","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/63d01f8087fb767073d48b55.docx"},{"id":96331074,"identity":"94496ad4-9f24-420a-b862-ac87cb841551","added_by":"auto","created_at":"2025-11-20 00:53:37","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":605506,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryinformationTRIPODAIchecklist.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7981034/v1/8cac7a17edac5d08707b3573.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Fusion of Intratumoral and Peritumoral Ultrasound Radiomics with Clinical Features for Distinguishing Diffuse Large B-cell Lymphoma from Hodgkin Lymphoma","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eLymphoma represents a major public health challenge, accounting for a substantial number of new hematologic malignancies diagnosed each year[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. This group of diseases is broadly divided into Hodgkin lymphoma and non-Hodgkin lymphoma, both of which display considerable heterogeneity in clinical behavior and pathological presentation[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Treatment selection and patient outcomes vary significantly based on the lymphoma subtype, molecular profile, and disease stage at diagnosis[\u003cspan additionalcitationids=\"CR4 CR5\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eCurrently, pathological biopsy remains the gold standard for subclassifying lymphomas. The choice of biopsy method is influenced by the suspected subtype and the clinical context. Ultrasound-guided core needle biopsy or fine-needle aspiration, owing to their minimally invasive nature, have become widely used diagnostic approaches for non-Hodgkin lymphoma with high tumor cellularity, particularly DLBCL.However, in HL, the sparse distribution of diagnostic Reed-Sternberg cells within a complex stroma often yields low diagnostic accuracy, frequently requiring a repeat biopsy[\u003cspan additionalcitationids=\"CR8\" citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Moreover, even with adequate tissue sampling, pathological overlap between certain lymphoma subtypes can pose diagnostic challenges.\u003c/p\u003e\u003cp\u003eClinically, lymphoma most commonly manifests as enlarged superficial lymph nodes, particularly in the cervical region. Owing to its non-invasive nature and high patient acceptability, ultrasonography is routinely used as the initial imaging technique in the diagnostic workup of lymphoma. However, conventional ultrasound alone cannot reliably distinguish between different lymphoma subtypes.\u003c/p\u003e\u003cp\u003eIn recent years, radiomics has drawn growing interest for its potential to enhance tumor diagnosis and enable precise treatment monitoring[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. This approach relies on the high-throughput extraction of quantitative features from medical images. By integrating these features with machine learning or deep learning models, radiomics can uncover valuable information related to disease diagnosis, prognosis, and therapeutic response. As an innovative tool, it helps overcome the subjectivity inherent in conventional image interpretation, promoting a shift from qualitative assessment to quantitative analysis, and has established itself as a useful non-invasive adjunct in precision oncology[\u003cspan additionalcitationids=\"CR13\" citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eWhile radiomic applications in CT, \u0026sup1;⁸F-FDG PET/CT, and MRI have shown significant potential for lymphoma subtyping[\u003cspan additionalcitationids=\"CR16 CR17\" citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e],ultrasound-based radiomics remains underexplored.. Furthermore, the diagnostic relevance of the peritumoral region, which reflects key tumor properties, has gained increasing recognition[\u003cspan additionalcitationids=\"CR20 CR21\" citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e].Therefore, this study aimed to integrate clinical features with intratumoral and peritumoral ultrasound radiomics features to distinguish DLBCL from HL. We present this article in accordance with the TRIPODAI checklist[\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e].\u003c/p\u003e"},{"header":"2. Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e2.1.Study design and patient population\u003c/h2\u003e\u003cp\u003e The study protocol received approval from the Medical Ethics Committee of The First Affiliated Hospital of Chongqing Medical University (Approval Number: 2024-364-01), and all methods were conducted in accordance with the relevant guidelines and regulations, including the Declaration of Helsinki. The requirement for informed consent was waived by the ethics committee due to the retrospective nature of the study.We retrospectively collected clinical and imaging data of patients with lymphoma admitted to The First Affiliated Hospital of Chongqing Medical University between January 2019 and January2025. The patient inclusion and exclusion criteria, along with the recruitment process, are presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. In total, 102 DLBCL patients with a total of 218 enlarged lymph nodes and 37 HL patients with a total of 105 enlarged lymph nodes were included in the study,and 323 enlarged lymph nodes were randomly allocated to the training (n\u0026thinsp;=\u0026thinsp;227) or test (n\u0026thinsp;=\u0026thinsp;96) set at a ratio of 7:3. A flowchart of the study is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e2.2.US Examination Methods and Image Acquisition\u003c/h2\u003e\u003cp\u003eIn this study, all images were acquired by three senior physicians with more than 10 years of experience in ultrasound diagnostics. Examination was conducted using a TOSHIBA Aplio 500 with a 14L5 broadband linear array probe (5\u0026ndash;14 MHz) or a Mindray Resona R9T equipped with an L14-5WU wideband linear array probe (5\u0026ndash;14 MHz).The ultrasound properties of the lymph nodes were observed using dynamic B-US scanning, and the longest-axis view of each lymph node was collected. Finally, US pictures were output using DICOM format.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\u003ch2\u003e2.3.Acquisition of clinical data and evaluation of US images\u003c/h2\u003e\u003cp\u003eClinical data, including patient age, sex, lactate dehydrogenase (LDH) levels, C-reactive protein (CRP) levels, hemoglobin (HB) levels, B symptoms, and Ann Arbor stage, were collected from the hospital information system. Lymph node characteristics recorded from B-mode ultrasound included: location (left cervical, right cervical, left axillary, right axillary, left inguinal, and right inguinal); border (well-defined, ill-defined, or indistinct); shape (regular, mildly irregular, or irregular); grouping (absent or present); cortical echogenicity (hypoechoic non-differentiated, markedly hypoechoic non-differentiated, hypoechoic differentiated, markedly hypoechoic differentiated, or isoechoic non-differentiated); internal echogenicity (isoechoic, hypoechoic, or markedly hypoechoic); posterior acoustic enhancement (present or absent); perinodal hyperechogenic ring (present or absent); hilum (present or absent); nodal-core sign (present, partially present, or absent); reticulation (present or absent); cystic degeneration (present or absent); calcification (present or absent); long-axis diameter; short-axis diameter; and long-to-short axis ratio. For color Doppler ultrasound, blood vessel architecture (avascular, transcapsular, combined, or hilar) and VITA\u0026trade; color score (ranging from 0 to 4) were recorded[\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eAmong 139 eligible patients, 918 ultrasound images were initially screened. Multiple enlarged lymph nodes (\u0026le;\u0026thinsp;6 per patient) were chosen for delineation. The selection process was conducted by an experienced physician (Physician A), who evaluated all images based on the following criteria: (1) complete visualization of both the tumor region and surrounding structures for comprehensive radiomics analysis; and (2) enlarged lymph nodes with a long-axis diameter\u0026thinsp;\u0026ge;\u0026thinsp;1 cm.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003e2.4.Rigion of Interest (ROI) Segmentation\u003c/h2\u003e\u003cp\u003eROI delineation was performed semi-automatically on the artificial intelligence research platform DARWIN (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.yizhun-ai.com\u003c/span\u003e\u003cspan address=\"http://www.yizhun-ai.com\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Initially, Physician A, who has more than three years of lymphoma diagnostic experience, annotated 100 randomly selected cases for training purposes under the supervision of senior Physician B, who has 15 years of clinical experience. Subsequently, Physician A independently annotated the remaining ROIs. Based on these ROIs, the platform automatically generated peritumoral regions of 8-, 16-, and 24-pixel widths using the seed region growing algorithm. Critically, all annotations were performed without knowledge of the pathological results.\u003c/p\u003e\u003cp\u003eTo evaluate the reproducibility of the ROI delineation and subsequent feature extraction, 50 cases were randomly selected for re-evaluation after a 20-day interval. Physician A redelineated the ROIs for these cases following the original protocol, blinded to the initial segmentations. Only radiomic features demonstrating excellent reproducibility, defined as an intraclass correlation coefficient greater than 0.8, were retained for further analysis.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\u003ch2\u003e2.5.Radiomics Feature Extraction and Radiomics Model Construction\u003c/h2\u003e\u003cp\u003eThe open-source Python package PyRadiomics (version 1.3.0) has been utilized to extract intratumoral and peritumoral radiomic features. A total of 1125 radiomic features were extracted, including 9 shape-based features, 18 first-order statistical features, 75 texture features, and 1023 transform-filtered features[\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eTo mitigate class imbalance and select informative features, we first applied the Synthetic Minority Over Sampling Technique (SMOTE; k\u0026thinsp;=\u0026thinsp;5) to the training set, where the HL class was oversampled by generating synthetic instances through feature space interpolation rather than direct replication[\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. Following resampling, all features were standardized using Z-score normalization. Feature selection was then performed in three sequential steps: (1) univariate filtering based on the ANOVA F-test (retaining features with \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05); (2) dimensionality reduction via the minimum Redundancy Maximum Relevance (mRMR) algorithm; and (3) refinement using elastic-net regularized logistic regression (α\u0026thinsp;=\u0026thinsp;0.5, mixing L1 and L2 penalties). The regularization parameter λ was optimized via 10-fold cross-validation, and only features with non-zero coefficients in the final model were retained for downstream analysis.\u003c/p\u003e\u003cp\u003eWe used the Logistic Regression Classifier to construct predictive models in the training group. Ten-fold cross-validation was employed on the training data set to assess the model\u0026rsquo;s predictive performance. The diagnostic accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the models with no (Rad), 8-pixel (Radexpand8), 16-pixel (Radexpand16), and 24-pixel (Radexpand24) outward extension were assessed using the confusion matrix and corresponding diagnostic metrics. AUC and DCA were used to assess the predictive performance and clinical value of the models. Calibration curves were used to evaluate the goodness-of-fit of the models. DeLong\u0026rsquo;s test was used to compare the differences in efficiency between the models.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003e2.6.Clinical Model and Fusion Models\u003c/h2\u003e\u003cp\u003eTo develop the clinical model, we first balanced using SMOTE (k\u0026thinsp;=\u0026thinsp;5), followed by Z-score normalization of all features. Subsequently, an ANOVA F-test was performed on clinical indicators and ultrasound characteristics. Variables that remained statistically significant (P\u0026thinsp;\u0026lt;\u0026thinsp;0.05) in the training set were subsequently incorporated into a multivariate logistic regression model.\u003c/p\u003e\u003cp\u003eBased on the optimal model among the four radiomics models,two fusion strategies were adopted to construct two fusion models. The first was a feature-level fusion model, constructed by concatenating all radiomics and clinical features into a single vector. This combined feature set underwent SMOTE (k\u0026thinsp;=\u0026thinsp;5) oversampling and Z-score normalization, followed by a three-step feature selection process consisting of F-test, mRMR, and LASSO. A logistic regression classifier was then trained on the finalized features, following the same methodology used in previous radiomics analyses.\u003c/p\u003e\u003cp\u003eThe second strategy employed decision-level fusion, which integrated the Radscore from the optimal radiomics model with statistically significant clinical and ultrasound characteristics. After similar preprocessing with SMOTE and Z-score normalization, a logistic regression classifier was trained to build the fusion model. Both fusion models were comprehensively evaluated based on discrimination performance including AUC, accuracy, sensitivity and specificity, calibration using calibration curves, and clinical usefulness via decision curve analysis.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003e2.7.Statistical analysis\u003c/h2\u003e\u003cp\u003eThe statistical analyses and data visualization were conducted using Python 3.969 and R 4.5.0. Continuous variables were summarized as median with interquartile range or mean with standard deviation, as appropriate.Normality and homogeneity of variance tests were performed on the quantitative data. For data that followed a normal distribution, independent samples t-test was used. For continuous variables that did not follow a normal distribution, the Mann-Whitney U test was applied. For categorical variables, either Fisher\u0026rsquo;s exact test or the chi-square test was used. Statistical significance was set at a two tailed \u003cem\u003eP\u003c/em\u003e value\u0026thinsp;\u0026lt;\u0026thinsp;0.05.\u003c/p\u003e\u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003e3.1. Patient Baseline Characteristics\u003c/h2\u003e\u003cp\u003eThis study enrolled 102 patients with DLBCL and 37 with HL, comprising 62 males and 77 females. The cohort had a median age of 61.5 years ranging from 14 to 97 years. Statistical analysis revealed that age, LDH class and LDH levels were significant factors (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05). In contrast, no significant differences were observed in sex, involvement stage, B symptoms, CRP class, CRP, HB class and HB (all \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05; Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eThe baseline characteristics of patients with DLBCL or HL\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"4\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCharacteristics\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDLBCL(n\u0026thinsp;=\u0026thinsp;102)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eHL (n\u0026thinsp;=\u0026thinsp;37)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSex\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.247\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eFemale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e60(58.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e17(45.9)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e42(41.2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e20(54.1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge (years)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e62.8(\u0026plusmn;\u0026thinsp;12.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e48.3(\u0026plusmn;\u0026thinsp;18.9)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e\u0026lt;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eInvolvement stage\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.490\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eII\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e24 (23.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e10(27.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eIII\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e20 (19.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e10(27.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eIV\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e58(56.9)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e17(46.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eB symptoms\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.554\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e42(41.2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e18(48.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNO\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e60(58.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e19 (51.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLDH class\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.020\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNormal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e51 (50.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e25 (67.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAbnormal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e51 (50.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e12 (32.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLDH (U/L)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e245.5 (186.0\u0026minus;480.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e200.0 (175.0\u0026minus;258.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.012\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCRP class\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.917\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNormal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e22(21.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e7(18.9)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAbnormal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e80(78.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e30(81.1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCRP(mg/L)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e28.6 (11.3\u0026ndash;58.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e39.0 (11.3\u0026ndash;81.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.267\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHB class\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.898\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNormal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e41(40.2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e16(43.2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAbnormal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e61(59.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e21(56.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHB(g/L)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e111.5 (98.0-122.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e110.0 (95.0-122.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.739\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003e3.2.Ultrasound image feature analysis\u003c/h2\u003e\u003cp\u003eStatistical analysis revealed that several ultrasonographic features differed significantly between DLBCL and HL. In the training group, these features were lymph node shape, grouping, cortical echogenicity, internal echogenicity, and posterior acoustic enhancement (all \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05). In the test group, significant differences were identified in lymph node shape, internal echogenicity and posterior acoustic enhancement (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05). In contrast, the distribution of all other US features did not reach statistical significance in either group (all \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05), as detailed in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eThe US features of patients with DLBCL or HL\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"8\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e\u003cp\u003eTraining group\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colspan=\"3\" nameend=\"c8\" namest=\"c6\"\u003e\u003cp\u003eTest group\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDLBCL(n\u0026thinsp;=\u0026thinsp;152)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eHL(n\u0026thinsp;=\u0026thinsp;75)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eDLBC(n\u0026thinsp;=\u0026thinsp;66)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003eHL(n\u0026thinsp;=\u0026thinsp;30)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c8\"\u003e\u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLocation(%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.281\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.351\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eright cervical\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e41(26.97)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e23(30.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e20(30.30)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e13(43.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eleft cervical\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e52(34.21)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e34(45.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e24(36.36)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e12(40.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eright axillary\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e15(9.87)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e6(8.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e6(9.09)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0(0.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eleft axillary\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e15(9.87)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e6(8.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e5(7.58)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e3(10.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eright inguinal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e14(9.21)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e3(4.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e8(12.12)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e1(3.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eleft inguinal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e15(9.87)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e3(4.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e3(4.55)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e1(3.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBorder(%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.094\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.089\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ewell-defined\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e150(98.68)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e71(94.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e65(98.48)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e27(90.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eill-defined\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e2(1.32)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e4(5.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e1(1.52)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e3(10.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eShape(%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.003\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eregular\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e136(89.47)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e54 (72.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e57(86.36)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e16(53.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003emildly irregular\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e14(9.21)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e17(22.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e9(13.64)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e14(46.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eirregular\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e2(1.32)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e4(5.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0 (0.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0 (0.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGrouping(%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.027\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.797\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eabsent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e109(71.71)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e42(56.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e43(65.15)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e18(60.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003epresent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e43(28.29)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e33(44.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e23(34.85)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e12(40.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCortical(%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.018\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.364\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ehypoechoic non-differentiated\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e115(75.66)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e61 (81.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e52(78.79)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e24(90.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003emarkedly hypoechoic non-differentiated\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e12(7.89)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e8 (10.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e6(9.09)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e2(6.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ehypoechoic differentiated\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e25(16.45)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e4 (5.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e8(12.12)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e1(3.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003emarkedly hypoechoic differentiated\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0(0.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e2 (2.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0 (0.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0(0.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eInternal(%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.013\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eisoechoic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e31(19.74)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e3(5.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e8(12.12)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0(0.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ehypoechoic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e106(69.74)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e71(92.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e51(77.27)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e30(100.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003emarkedly hypoechoic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e16(10.53)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0(2.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e7(10.61)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0(0.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePosterior(%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003epresent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e30(19.74)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e63(84.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e11(16.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e24(80.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eabsent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e122(80.26)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e12(16.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e55(83.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e6(20.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRing(%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.254\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.587\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003epresent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1(0.66)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e2( 2.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e2(3.03)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e2(6.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eabsent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e151(99.34)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e73(97.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e64(96.92)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e28(93.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHilum(%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.315\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.770\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003epresent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e39(25.66)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e14(18.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e11(16.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e4(13.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eabsent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e113(74.34)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e61(81.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e55(83.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e26(86.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNodal-core sign(%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.338\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.522\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ecomplete\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e20(13.16)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e5(6.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e8(12.12)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e2(6.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003epartial\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e22(14.47)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e11(16.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e8(12.12)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e2(6.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eabsent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e110(72.37)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e58(77.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e50(75.76)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e26(86.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eReticulation(%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.161\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.229\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003epresent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e4(2.63)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e5(6.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e1(1.52)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e2(6.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eabsent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e148(97.37)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e70(93.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e65(98.48)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e28(93.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCystic(%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.222\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.172\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003epresent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e8(5.26)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e8(10.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e6( 9.09)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0(0.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eabsent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e144(94.74)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e67(89.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e60(90.91)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e30(100.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCalcification(%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e1.000\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.074\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003epresent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e13(8.55)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e7(9.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e2( 3.03)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e4(13.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eabsent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e139(91.45)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e68(90.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e64(96.97)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e26 (86.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBlood-vessel architecture(%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.448\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.190\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eavascular\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e23(15.13)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e6(8.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e8(12.12)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e1(3.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003etranscapsular\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e93(61.18)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e52(69.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e45(68.18)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e24(80.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ecombined\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e30(19.74)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e15(20.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e12(18.18)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e3(10.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ehilar\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e6(3.95)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e2(2.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e1(1.52)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e2( 6.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eColor score(%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.126\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.500\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e23(14.38)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e6(8.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e8(12.12)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e1( 3.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e75(48.37)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e37(49.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e30(45.45)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e17(56.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e27(17.76)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e22(29.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e21(31.82)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e8(26.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e27(17.76)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e10(13.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e7(10.61)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e4( 13.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLong diameter(mm)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e25.00(18.23-31.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e27.00(18.80\u0026ndash;33.40)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.278\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e23.00(17.00\u0026ndash;33.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e24.90(18.00-28.23)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.785\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eShort diameter(mm)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e12.00 (9.00\u0026ndash;17.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e13.00 (10.00\u0026ndash;16.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.526\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e12.00(10.00\u0026ndash;17.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e13.00(11.2-14.38)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.596\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLong to Short(mm)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.95 (1.50\u0026ndash;2.36)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e2.00(1.65\u0026ndash;2.32)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.372\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e1.85(1.61\u0026ndash;2.12)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e1.84(1.52-2.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.323\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\u003ch2\u003e3.3.Clinical,US and Radiomics Feature Analysis\u003c/h2\u003e\u003cp\u003eA multivariate logistic regression analysis was performed incorporating the clinical indicators and conventional ultrasound features that were significant in the univariate analysis of the training group. A clinical model was derived from the results of this multivariate analysis to aid in differentiating DLBCL from HL. Furthermore, radiomics analysis extracted 10, 10, 9, and 8 key features from the intratumoral region and the peritumoral regions extended by 8, 16, and 24 pixels, respectively (Supplementary Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003e3.4.Clinical, Radiomics and Fusion Model Performance Analysis\u003c/h2\u003e\u003cp\u003eThe diagnostic indicators of the predictive models for each study set are presented in Tables\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e and \u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e. Using radiomics, the model with a 16-pixel outward extension (Radexpand16) achieved the highest AUC (0.636) for the test set (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Although DeLong\u0026rsquo;s test revealed no significant differences between the four radiomics models (all \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05), the calibration curve exhibited good consistency (Supplementary Figure \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e), and DCA demonstrated optimal clinical benefit (Supplementary Figure S3). In the fusion model, owing to the superior performance of the 16-pixel outward extension,we developed both featurelevel (Radexpand16_Clinical-F) and decision-level (Radexpand16_Clinical-R) fusion models.Two fusion models outperformed the Radexpand16 model on both the training and test sets (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The Radexpand16_Clinical-R model demonstrated the best performance, with an AUC of 0.941, accuracy of 0.876, sensitivity of 0.928, and specificity of 0.824 in the training set, and DeLong\u0026rsquo;s test showed significance (all \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05). The test set achieved an AUC of 0.896, accuracy of 0.887, sensitivity of 0.906, and specificity of 0.877,and it displayed optimal clinical benefit on the DCA (Supplementary Figure S3). Delong's test indicated that the AUC of the Radexpand16_Clinical_R model was statistically significantly compared to both the Rad model and the Clinical model (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05), while no statistically significant difference was observed when compared to the Radexpand16_Clinical_F model. To facilitate clinical application and intuitive understanding, we visualized the Radexpand16_Clinical_R model as a nomogram (Supplementary Figure S4).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eEvaluation results of four basic radiomics models in the training group and test group\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"11\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c11\" colnum=\"11\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eModels\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eGroup\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eAUC\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e95% CI\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eSensitivity\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eSpecificity\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003ePPV\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c8\"\u003e\u003cp\u003eNPV\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c9\"\u003e\u003cp\u003eF1 score\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c10\"\u003e\u003cp\u003eAccuracy\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRad\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003etraining\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.584\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.524\u0026ndash;0.644\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.852\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.341\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.564\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.698\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e\u003cp\u003e0.679\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e\u003cp\u003e0.597\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e0.130\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRadexpand8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003etraining\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.579\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.515\u0026ndash;0.643\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.588\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.582\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.584\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.586\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e\u003cp\u003e0.586\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e\u003cp\u003e0.585\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e0.115\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRadexpand16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003etraining\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.650\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.589\u0026ndash;0.712\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.771\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.497\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.605\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.685\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e\u003cp\u003e0.678\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e\u003cp\u003e0.634\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003eReference\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRadexpand24\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003etraining\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.698\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.639\u0026ndash;0.756\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.464\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.863\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.772\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.617\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e\u003cp\u003e0.580\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e\u003cp\u003e0.663\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e0.274\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRad\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003etest\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.531\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.409\u0026ndash;0.654\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.389\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.800\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.483\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.732\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e\u003cp\u003e0.431\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e\u003cp\u003e0.667\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e0.249\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRadexpand8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003etest\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.577\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.446\u0026ndash;0.708\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.594\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.662\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.463\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.768\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e\u003cp\u003e0.521\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e\u003cp\u003e0.639\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e0.528\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRadexpand16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003etest\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.636\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.509\u0026ndash;0.762\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.625\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.692\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.500\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.789\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e\u003cp\u003e0.556\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e\u003cp\u003e0.670\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003eReference\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRadexpand24\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003etest\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.581\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.461\u0026ndash;0.702\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.813\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.431\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.413\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.824\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e\u003cp\u003e0.547\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e\u003cp\u003e0.557\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e0.543\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"11\"\u003e\u003cem\u003eP\u003c/em\u003e values were calculated with the DeLong test.\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eEvaluation results of the best basic model,clinical model and fusion models in the training group and test group\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"12\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c11\" colnum=\"11\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c12\" colnum=\"12\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eModels\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eGroup\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eAUC\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e95% CI\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eSensitivity\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eSpecificity\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003ePPV\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c8\"\u003e\u003cp\u003eNPV\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c9\"\u003e\u003cp\u003eF1 score\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c10\"\u003e\u003cp\u003eAccuracy\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c11\"\u003e\u003cp\u003eP value\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c12\"\u003e\u003cp\u003eP value\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRadexpand16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003etraining\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.650\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.589\u0026ndash;0.712\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.771\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.497\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.605\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.685\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e\u003cp\u003e0.678\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e\u003cp\u003e0.634\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003eReference\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eClinical\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003etraining\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.521\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.452\u0026ndash;0.589\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.980\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.255\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.568\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.929\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e\u003cp\u003e0.719\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e\u003cp\u003e0.618\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e0.006\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRadexpand16_Clinical_R\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003etraining\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.941\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.916\u0026ndash;0.966\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.928\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.824\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.840\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.920\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e\u003cp\u003e0.882\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e\u003cp\u003e0.876\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u003cp\u003eReference\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRadexpand16_Clinical_F\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003etraining\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.935\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.907\u0026ndash;0.964\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.928\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.843\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.855\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.921\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e\u003cp\u003e0.890\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e\u003cp\u003e0.886\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u003cp\u003e0.676\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRadexpand16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003etest\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.636\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.509\u0026ndash;0.762\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.625\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.692\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.500\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.789\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e\u003cp\u003e0.556\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e\u003cp\u003e0.670\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003eReference\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eClinical\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003etest\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.552\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.436\u0026ndash;0.669\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.938\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.369\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.423\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.923\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e\u003cp\u003e0.583\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e\u003cp\u003e0.557\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e0.030\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRadexpand16_Clinica_R\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003etest\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.943\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.898\u0026ndash;0.989\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.906\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.877\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.784\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.950\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e\u003cp\u003e0.841\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e\u003cp\u003e0.887\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u003cp\u003eReference\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRadexpand16_Clinical_F\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003etest\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.938\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.890\u0026ndash;0.986\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.875\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.862\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.757\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.933\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e\u003cp\u003e0.812\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e\u003cp\u003e0.866\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u003cp\u003e0.852\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"12\"\u003e\u003cem\u003eP\u003c/em\u003e values were calculated with the DeLong test.\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003ch2\u003e3.5. Model Interpretation\u003c/h2\u003e\u003cp\u003eTo interpret the Radexpand16_Clinical_R model, feature importance analysis was employed to quantify the contribution of each feature. A bar plot was generated based on the absolute values of the model coefficients, representing the relative influence of each feature on the prediction outcome (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Notably, posterior acoustic enhancement emerged as the most stable and independent predictor for differential diagnosis, with higher values corresponding to an increased probability of DLBCL.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eIn this study, we established radiomics models based on ultrasound images of primary tumors and peritumoral images to differentiate DLBCL from HL.We developed two fusion models that integrated clinical data and radiomics features.The Radexpand16_Clinical_R model exhibited the best performance with an AUC of 0.943 for the test set.\u003c/p\u003e\u003cp\u003eFeature importance analysis proved valuable in interpreting the predictions of our fused Radexpand16_Clinical_R model. This model successfully integrated a combination of ultrasonographic features, radiomic features, and clinical parameters, with each category contributing complementary diagnostic information. The resulting feature profile aligns well with established pathological differences between the two lymphomas. Histologically, DLBCL consists of a monomorphic proliferation of large, rapidly dividing cells that are densely packed, creating a relatively uniform tissue architecture. We hypothesize that this homogeneity results in lower ultrasonic attenuation, providing a plausible explanation for the frequent observation of posterior acoustic enhancement in DLBCL. Conversely, the classic morphology of HL reveals a sparse population of Reed-Sternberg cells within a prominent background of inflammatory cells and fibrous stroma. This complex and heterogeneous microenvironment likely causes greater ultrasound beam attenuation, accounting for its different sonographic appearance[\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. At the serological level, the high metabolic activity and substantial tumor burden characteristic of DLBCL promote accelerated cellular turnover and destruction. This pathophysiology logically accounts for the elevated serum LDH levels frequently seen in DLBCL patients, reinforcing its role as a supportive biomarker in our model[\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eLogistic Regression is widely used in radiomics research as a machine learning model. As a generalized linear classifier, it applies a logistic function to convert a linear combination of input features into a probability score between 0 and 1, indicating the likelihood of the sample belonging to the target class. Binary classification is achieved by comparing this probability against a predefined threshold, typically 0.5. A key strength of logistic regression lies in the interpretability. The weight coefficients clearly quantify how much each feature contributes to the final prediction, which is important in clinical settings where model transparency is required. Furthermore, when working with limited medical imaging datasets, logistic regression tends to generalize better and is less prone to overfitting compared to deep learning approaches[\u003cspan additionalcitationids=\"CR32 CR33\" citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eTo explore the ability of peritumoral features to differentiate between DLBCL and HL, we compared the predictive performance of the models within expansion ranges of 8, 16 and 24pixels. In test group, the predictive performance within a 16-pixel expansion range was optimal, however, in training group, the model with a 16-pixel extension did not perform significantly better than other extension models, especially compared with 24-pixel expansion range.This may be due to limited valuable information in the peritumoral region, and the 16-pixel extension possibly including noise and irrelevant features[\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eIn this study, we attempted both feature-level and decision-level fusion model, and found that decision-level fusion model yielded the best results.This finding is in line with those of several prior researches in radiomics[\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e].However, some studies have reported better performance of feature-level fusion versus decision-level fusion[\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e].For example, Gan et al utilized Intratumoral and Peritumoral Endorectal Ultrasound Imaging based Deep Learning, Radiomics, and Fusion Models to Predict KRAS Gene Mutations in Rectal Cancer, and found that the feature-level fusion strategy yielded optimal performance[\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e].Therefore, it is necessary to develop a fusion model combining both approaches and select the optimal method.\u003c/p\u003e\u003cp\u003eThis study has several limitations. First, as a single-center, small sample retrospective study, it may suffer from selection bias and lacks an external validation cohort, making it difficult to assess the model's predictive performance accurately. Second, due to the limited sample size, we were unable to perform subgroup analyses based on lymphoma stage or aggressiveness, which constrains the broader clinical applicability of the results. Third, variations in operator experience and ultrasound equipment may have introduced heterogeneity in image quality, potentially influencing model development. Finally, the study utilized conventional ultrasound alone without integrating multimodal imaging data, such as contrast-enhanced ultrasound, CT, or MRI. Future studies should involve multicenter, large sample prospective research, include more stage IV or highly aggressive lymphoma cases for stratified analysis. Integrating multimodal imaging data could further improve the clinical value of radiomics models in guiding biopsy procedures and enabling noninvasive subtype classification.\u003c/p\u003e"},{"header":"5. Conclusions","content":"\u003cp\u003eOur study presents a non-invasive preoperative method based on radiomics models to differentiate between diffuse large B-cell lymphoma and Hodgkin lymphoma. This approach provides valuable support for clinical diagnosis and aids in optimizing biopsy strategies and treatment decisions. The Radexpand16_Clinical_R model, developed using pre-biopsy 2D grayscale ultrasound images, effectively discriminates between these two lymphoma subtypes. Incorporating peritumoral regions further improved its predictive performance. For patients with a high likelihood of HL, clinicians are advised to prioritize excisional biopsy and perform targeted immunohistochemical testing.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eFunding\u0026nbsp;\u003c/strong\u003eThe authors received no funding for this work.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u0026nbsp; The authors have no competing interests to declare.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability\u003c/strong\u003e\u0026nbsp; The data are available from the corresponding author upon reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u0026nbsp;\u003c/strong\u003eThe authors would like to thank the clinicians and technicians in the Department of Ultrasound for their assistance in data acquisition and management. We are also grateful to the team behind the DARWIN AI research platform for their technical support. Our deepest gratitude goes to all the patients who participated in this study. Finally, we thank our families for their unwavering support throughout this project.\u003c/p\u003e\n"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eSiegel RL, Kratzer TB, Giaquinto AN, et al (2025) Cancer statistics, 2025. CA Cancer J Clin 75:10\u0026ndash;45. https://doi.org/10.3322/caac.21871\u003c/li\u003e\n\u003cli\u003eCheson BD, Fisher RI, Barrington SF, et al (2014) Recommendations for initial evaluation, staging, and response assessment of Hodgkin and non-Hodgkin lymphoma: the Lugano classification. J Clin Oncol 32:3059\u0026ndash;3068. https://doi.org/10.1200/JCO.2013.54.8800\u003c/li\u003e\n\u003cli\u003eRossi C, Manson G, Marouf A, et al (2024) Classic Hodgkin Lymphoma: The LYSA pragmatic guidelines. Eur J Cancer 213:115073. https://doi.org/10.1016/j.ejca.2024.115073\u003c/li\u003e\n\u003cli\u003eAlaggio R, Amador C, Anagnostopoulos I, et al (2022) The 5th edition of the World Health Organization Classification of Haematolymphoid Tumours: Lymphoid Neoplasms. Leukemia 36:1720\u0026ndash;1748. https://doi.org/10.1038/s41375-022-01620-2\u003c/li\u003e\n\u003cli\u003eBarraclough A, Hawkes E, Sehn LH, Smith SM (2024) Diffuse large B-cell lymphoma. Hematological Oncology 42:e3202. https://doi.org/10.1002/hon.3202\u003c/li\u003e\n\u003cli\u003eBerhan A, Almaw A, Damtie S, Solomon Y (2025) Diffuse large B cell lymphoma (DLBCL): epidemiology, pathophysiology, risk stratification, advancement in diagnostic approaches and prospects: narrative review. Discov Onc 16:184. https://doi.org/10.1007/s12672-025-01958-w\u003c/li\u003e\n\u003cli\u003eK\u0026uuml;ppers R (2009) The biology of Hodgkin\u0026rsquo;s lymphoma. Nat Rev Cancer 9:15\u0026ndash;27. https://doi.org/10.1038/nrc2542\u003c/li\u003e\n\u003cli\u003eKwon Y, Lee MK (2025) Diagnostic Performance and Safety of Ultrasound-Guided Core Needle Biopsy for Diagnosing Lymphoma: A Systematic Review and Meta-Analysis. Cancer Medicine 14:e70414. https://doi.org/10.1002/cam4.70414\u003c/li\u003e\n\u003cli\u003eHoppe RT, Advani RH, Ai WZ, et al (2020) Hodgkin Lymphoma, Version 2.2020, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw 18:755\u0026ndash;781. https://doi.org/10.6004/jnccn.2020.0026\u003c/li\u003e\n\u003cli\u003eCamps J, Jim\u0026eacute;nez-Franco A, Garc\u0026iacute;a-Pablo R, et al (2025) Artificial intelligence-driven integration of multi-omics and radiomics: A new hope for precision cancer diagnosis and prognosis. Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease 1871:167841. https://doi.org/10.1016/j.bbadis.2025.167841\u003c/li\u003e\n\u003cli\u003eLin P, Zheng J, Liu C, et al (2025) Radiogenomic insights suggest that multiscale tumor heterogeneity is associated with interpretable radiomic features and outcomes in cancer patients. Computerized Medical Imaging and Graphics 124:102586. https://doi.org/10.1016/j.compmedimag.2025.102586\u003c/li\u003e\n\u003cli\u003eLambin P, Leijenaar RTH, Deist TM, et al (2017) Radiomics: the bridge between medical imaging and personalized medicine. Nat Rev Clin Oncol 14:749\u0026ndash;762. https://doi.org/10.1038/nrclinonc.2017.141\u003c/li\u003e\n\u003cli\u003eRogers W, Thulasi Seetha S, Refaee TAG, et al (2020) Radiomics: from qualitative to quantitative imaging. Br J Radiol 93:. https://doi.org/10.1259/bjr.20190948\u003c/li\u003e\n\u003cli\u003eQi Y-J, Su G-H, You C, et al (2024) Radiomics in breast cancer: Current advances and future directions. Cell Rep Med 5:101719. https://doi.org/10.1016/j.xcrm.2024.101719\u003c/li\u003e\n\u003cli\u003eSi J, Wang H, Xie M, et al (2024) Distinguishing diffuse large B-cell lymphoma from Hodgkin\u0026rsquo;s lymphoma in children using an enhanced computed tomography radiomics approach. Transl Pediatr 13:716\u0026ndash;726. https://doi.org/10.21037/tp-23-586\u003c/li\u003e\n\u003cli\u003eChauvie S, Kostakoglu Shields L, Guerra L, et al (2024) PET Radiomics Discriminate Among Different Lymphoma Sub-Types. Blood 144:7505. https://doi.org/10.1182/blood-2024-205170\u003c/li\u003e\n\u003cli\u003eSi J, Wang H, Xie M, et al (2024) The value of radiomics features of the spleen as surrogates for differentiating subtypes of common pediatric lymphomas. QUANTITATIVE IMAGING IN MEDICINE AND SURGERY 14:. https://doi.org/10.21037/qims-24-122\u003c/li\u003e\n\u003cli\u003eDuan C ling, An L, Yang Y feng, et al (2025) The Role of Artificial Intelligence and Radiomics in the Management of Lymphomas by PET/CT: The Clairvoyance in Clinic. Cancer Management and Research 17:1457\u0026ndash;1475. https://doi.org/10.2147/CMAR.S529589\u003c/li\u003e\n\u003cli\u003eSun J, Wu X, Ma R, et al (2025) Intratumoral and peritumoral radiomics of MRI predicts pathologic complete response to neoadjuvant chemoimmunotherapy in patients with head and neck squamous cell carcinoma. J Immunother Cancer 13:e011189. https://doi.org/10.1136/jitc-2024-011189\u003c/li\u003e\n\u003cli\u003eHong J, Yang L, Huo J, et al (2025) Predicting the invasiveness of pulmonary adenocarcinoma using intratumoral and peritumoral radiomics features. Front Med 12:. https://doi.org/10.3389/fmed.2025.1541682\u003c/li\u003e\n\u003cli\u003eXu H, Yang A, Kang M, et al (2025) Intratumoral and peritumoral radiomics signature based on DCE-MRI can distinguish between luminal and non-luminal breast cancer molecular subtypes. Sci Rep 15:14720. https://doi.org/10.1038/s41598-025-98155-0\u003c/li\u003e\n\u003cli\u003eHuang Z, Mo S, Wu H, et al (2024) Optimizing breast cancer diagnosis with photoacoustic imaging: An analysis of intratumoral and peritumoral radiomics. Photoacoustics 38:100606. https://doi.org/10.1016/j.pacs.2024.100606\u003c/li\u003e\n\u003cli\u003eMoons KGM, Altman DG, Reitsma JB, et al (2015) Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD): explanation and elaboration. Ann Intern Med 162:W1-73. https://doi.org/10.7326/M14-0698\u003c/li\u003e\n\u003cli\u003eFischerova D, Garganese G, Reina H, et al (2021) Terms, definitions and measurements to describe sonographic features of lymph nodes: consensus opinion from the Vulvar International Tumor Analysis (VITA) group. Ultrasound Obstet Gynecol 57:861\u0026ndash;879. https://doi.org/10.1002/uog.23617\u003c/li\u003e\n\u003cli\u003eAbbasian Ardakani A, Bureau NJ, Ciaccio EJ, Acharya UR (2022) Interpretation of radiomics features\u0026ndash;A pictorial review. Computer Methods and Programs in Biomedicine 215:106609. https://doi.org/10.1016/j.cmpb.2021.106609\u003c/li\u003e\n\u003cli\u003eSafder MU, Naveed SS, Khurshid K, et al (2025) Optimizing imbalanced learning with genetic algorithm. Sci Rep 15:34857. https://doi.org/10.1038/s41598-025-09424-x\u003c/li\u003e\n\u003cli\u003eSatou A, Takahara T, Nakamura S (2022) An Update on the Pathology and Molecular Features of Hodgkin Lymphoma. Cancers 14:2647. https://doi.org/10.3390/cancers14112647\u003c/li\u003e\n\u003cli\u003eLi S, Young KH, Medeiros LJ (2018) Diffuse large B-cell lymphoma. Pathology 50:74\u0026ndash;87. https://doi.org/10.1016/j.pathol.2017.09.006\u003c/li\u003e\n\u003cli\u003eZhang J, Lu Q, Liu W, Zhou N (2025) The metabolic role of lactate dehydrogenase in the growth of diffuse large B cell lymphoma. Ann Hematol 104:545\u0026ndash;558. https://doi.org/10.1007/s00277-024-06083-2\u003c/li\u003e\n\u003cli\u003eZhang J, Chen B, Zhang C, et al (2024) The prognostic value of hydroxybutyrate dehydrogenase in diffuse large B cell lymphoma. iScience 27:. https://doi.org/10.1016/j.isci.2024.110905\u003c/li\u003e\n\u003cli\u003eZabor EC, Reddy CA, Tendulkar RD, Patil S (2022) Logistic Regression in Clinical Studies. International Journal of Radiation Oncology*Biology*Physics 112:271\u0026ndash;277. https://doi.org/10.1016/j.ijrobp.2021.08.007\u003c/li\u003e\n\u003cli\u003eShi S, An X, Li Y (2023) Ultrasound Radiomics-Based Logistic Regression Model to Differentiate Between Benign and Malignant Breast Nodules. J Ultrasound Med 42:869\u0026ndash;879. https://doi.org/10.1002/jum.16078\u003c/li\u003e\n\u003cli\u003eSuttaket T, Vardhan LVH, Kok S (2024) Interpretable Predictive Models for Healthcare via Rational Logistic Regression\u003c/li\u003e\n\u003cli\u003eDemircioğlu A (2023) Are deep models in radiomics performing better than generic models? A systematic review. European Radiology Experimental 7:11. https://doi.org/10.1186/s41747-023-00325-0\u003c/li\u003e\n\u003cli\u003eHuang L, Xu L, Wang X, et al (2025) Prediction of EGFR Mutations in Lung Adenocarcinoma via CT Images: A Comparative Study of Intratumoral and Peritumoral Radiomics, Deep Learning, and Fusion Models. Academic Radiology 32:4880\u0026ndash;4892. https://doi.org/10.1016/j.acra.2025.04.029\u003c/li\u003e\n\u003cli\u003eGan Y, Hu Q, Shen Q, et al (2025) Comparison of Intratumoral and Peritumoral Deep Learning, Radiomics, and Fusion Models for Predicting KRAS Gene Mutations in Rectal Cancer Based on Endorectal Ultrasound Imaging. Ann Surg Oncol 32:3019\u0026ndash;3030. https://doi.org/10.1245/s10434-024-16697-5\u003c/li\u003e\n\u003cli\u003eZhang W, Zhao X, Meng L, et al (2025) A Multicentre Comparative Analysis of Radiomics, Deep-learning, and Fusion Models for Predicting Postpartum Hemorrhage. Academic Radiology. https://doi.org/10.1016/j.acra.2025.05.068\u003c/li\u003e\n\u003cli\u003eLi Z, Qin Y, Liao X, et al (2025) Comparison of clinical, radiomics, deep learning, and fusion models for predicting early recurrence in locally advanced rectal cancer based on multiparametric MRI: a multicenter study. European Journal of Radiology 189:112173. https://doi.org/10.1016/j.ejrad.2025.112173\u003c/li\u003e\n\u003cli\u003eLi X, Yang L, Jiao X (2023) Comparison of Traditional Radiomics, Deep Learning Radiomics and Fusion Methods for Axillary Lymph Node Metastasis Prediction in Breast Cancer. Academic Radiology 30:1281\u0026ndash;1287. https://doi.org/10.1016/j.acra.2022.10.015\u003c/li\u003e\n\u003cli\u003eMohsen F, Ali H, El Hajj N, Shah Z (2022) Artificial intelligence-based methods for fusion of electronic health records and imaging data. Sci Rep 12:17981. https://doi.org/10.1038/s41598-022-22514-4\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Radiomics, Ultrasonography, Diffuse large B-cell Lymphom, Hodgkin Lymphoma","lastPublishedDoi":"10.21203/rs.3.rs-7981034/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7981034/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eObjectives\u003c/h2\u003e\u003cp\u003eTo develop a model integrating clinical features with intratumoral and peritumoral ultrasound radiomics features for distinguishing Diffuse Large B-Cell Lymphoma (DLBCL) from Hodgkin Lymphoma (HL).\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e\u003cp\u003eThis study retrospectively analyzed data from 102 DLBCL patients (218 enlarged lymph nodes) and 37 HL patients (105 enlarged lymph nodes) at The First Affiliated Hospital of Chongqing Medical University. The enlarged lymph nodes were randomly allocated to a training cohort (227 nodes) and a test cohort (96 nodes) in a 7:3 ratio. Radiomics models were constructed based on ultrasound images from both the primary tumor and the peritumoral area. For the optimally performing regions-of-interest, two fusion strategies, a feature-based and a decision-based model, were applied to build the fusion models. The performance of each model was assessed using the area under the receiver operating characteristic curve (AUC) and decision curve analysis (DCA). The study strictly adhered to the TRIPODAI checklist.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e\u003cp\u003eIn the test cohort, the radiomics model using a 16-pixel peritumoral extension demonstrated optimal performance with an AUC of 0.636. The decision-based fusion model (Radexpand16_Clinica_R) achieved the highest AUC (0.943) among all models evaluated. Additionally, the Radexpand16_Clinica_R model exhibited excellent sensitivity, specificity, and favorable clinical utility as shown by DCA.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e\u003cp\u003eThe decision-based fusion model Radexpand16_Clinica_R effectively differentiates DLBCL from HL. Incorporating peritumoral regions improved the predictive capability of the radiomics models.\u003c/p\u003e","manuscriptTitle":"Fusion of Intratumoral and Peritumoral Ultrasound Radiomics with Clinical Features for Distinguishing Diffuse Large B-cell Lymphoma from Hodgkin Lymphoma","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-11-20 00:53:23","doi":"10.21203/rs.3.rs-7981034/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"307f6e76-0b98-44bf-9939-530772d2d564","owner":[],"postedDate":"November 20th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":58151882,"name":"Biological sciences/Cancer"},{"id":58151883,"name":"Health sciences/Oncology"}],"tags":[],"updatedAt":"2026-01-21T11:58:49+00:00","versionOfRecord":[],"versionCreatedAt":"2025-11-20 00:53:23","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7981034","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7981034","identity":"rs-7981034","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}