Kinetic heterogeneity is associated with axillary lymph node metastasis in cN0 breast cancer based on dynamic contrast-enhanced magnetic resonance imaging radiomics nomogram

Kinetic heterogeneity is associated with axillary lymph node metastasis in cN0 breast cancer based on dynamic contrast-enhanced magnetic resonance imaging radiomics nomogram | 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 Research Article Kinetic heterogeneity is associated with axillary lymph node metastasis in cN0 breast cancer based on dynamic contrast-enhanced magnetic resonance imaging radiomics nomogram Tongxu Shen, Dingli Ye, Ming Yao, Jieqiong Yan, Han Zhang, Shuangyan Sun This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4445164/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 Background To investigate whether kinetic heterogeneity, assessed via dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI)-based radiomics nomogram, associated with axillary lymph node metastasis (ALNM) in cN0 breast cancer. Methods 373 consecutive women with cN0 breast cancer who underwent preoperative DCE-MRI were retrospectively evaluated from 2016 to 2020. The kinetic heterogeneity (a measure of heterogeneity in the proportions of peak enhancement, peak enhancement ratio, persistent, plateau, and washout) was assessed with DCE-MRI using B.K. software automatically. Radiomics features were extracted from magnetic resonance imaging (MRI) images of the primary breast cancer lesion. The minimum redundancy maximum relevance algorithm was used to select ALNM positively-related features and radiomics score was constructed. Clinical features, MRI features, kinetic heterogeneity, and radiomics score were screened out by multivariate logistic regression analysis, and the nomogram was constructed from these characteristics. Possible associations between DCE-MRI-based kinetic heterogeneity and ALNM were analyzed. The unsupervised clustering K-Mean algorithm was use to risk stratification. Results Five independent risk factors were screened out to build the nomogram, including: age, margin, ratio, washout, and radiomics score. The area under the receiver operating characteristic curve was 0.857 and 0.858 in the training and test cohorts, respectively. The risk stratification system divided all patients into three risk groups. Axillary lymph node dissection was not recommended for the low-risk group and was strongly recommended for the high-risk group. Conclusions Radiomic analysis of kinetic heterogeneity based on the DCE-MRI images has the potential to more accurately identify tumor kinetic features and serve as a valuable clinical marker to enhance the prediction of ALNM in cN0 breast cancer. Kinetic heterogeneity axillary lymph node metastasis breast cancer dynamic contrast-enhanced magnetic resonance imaging radiomics nomogram Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction Breast cancer accounts for 30% of newly diagnosed female cancers, ranking second in the deaths from cancers in women, second only to lung cancer in 2021 [ 1 ]. The axillary lymph node (ALN) status affects the locoregional recurrence and overall survival rate [ 2 ]. Meanwhile, ALN status determines the scope of the surgery and the potential need for neoadjuvant chemotherapy. However, owing to the invasive nature of ALN dissection (ALND), the patients face considerable complications [ 3 , 4 ]. Therefore, preoperative noninvasive prediction of ALN metastasis (ALNM) is particularly important. Physical examination, mammography, ultrasound, and magnetic resonance imaging (MRI) are all commonly used for preoperative diagnosis of breast cancer; however, their ability to assess ALN is not satisfactory [ 4 ]. Because dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is independent of breast density, it is currently a sensitive and specific imaging technique for the diagnosis of breast cancer. It depicts the morphologic features of breast tumors and reveals dynamics of enhancement, which may reflect angiogenesis [ 5 ]. Breast cancer is highly heterogeneous in nature. The intra-tumoral heterogeneity is associated with molecular subtypes [ 6 , 7 ], metastasis [ 8 ], prognosis [ 5 ] and recurrence [ 9 ] in breast cancer, which may be due to intrinsic aggressive biology or treatment resistance. Radiomics transforms medical images into higher dimensional data, which has been widely used in clinical medicine. Prediction models are becoming recognized as valuable tools in recent years and are recommended in clinical practice guidelines [ 10 ]. Radiomics-based prediction model has been applied in the prediction of sentinel lymph node burden [ 3 ], human epidermal growth factor receptor 2 (HER 2) expression level [ 11 ], prognostic biomarkers and molecular subtypes [ 12 ] of breast cancer in the recent years. However, whether DCE-MRI radiomics-based nomogram could predict ALNM before surgery accurately, and whether it can be explained from the perspective of intra-tumoral heterogeneity, there are few relevant researches at present. We hypothesized that kinetic features of breast cancer and the heterogeneity of these features in preoperative DCE-MRI were associated with ALNM in women with breast cancer. The purposes of the study were, first, to extract the key elements for predicting ALNM, including clinical, DCE-MRI, and radiomics features. Second, to establish radiomics-based nomogram based on the key features for the prediction of ALNM. Third, to validate and evaluate the nomogram. Methods This retrospective study was approved by the institutional review board of Jilin Cancer Hospital (IRB No.2023-001-01). The requirement for informed consent was waive. Study population A retrospective review of medical records collected between 2016 and 2020 identified 2606 consecutive patients with breast cancer who underwent preoperative DCE-MRI. Patients were excluded if (a) male, (b) with incomplete clinical or pathological information, (c) with a history of other malignancy, (d) underwent excisional biopsy for diagnosis, (e) received neoadjuvant chemotherapy before surgery, (f) suspected of ALNM based on preoperative ultrasound or palpation, (g) with chemical shift artifacts or motion artifacts on the DCE-MRI images, (h) DCE-MRI examination time was over two weeks (Fig. 1 ). For multiple lesions, the lesion with the largest diameter was selected for subsequent data analysis. MRI acquisition and analysis MRI was performed in the prone position with a 3.0 T MRI system (Ingenia, Philips, Netherlands) using a dedicated eight-channel breast array coil, with the arms raised and the head advanced. Bilateral axial T2-weighted turbo spin-echo imaging and T1-weighted volumetric interpolated breath-hold examination with fat saturation were acquired. DCE-MRI images were acquired using a three-dimensional fat suppressed T1-weighted sequence and included one pre-contrast and five post-contrast images (repetition time, 4.6 ms; echo time, 2.0 ms; flip angle, 12°; field of view, 300×380 mm 2 ; matrix, 300×380; section thickness, 1 mm; no gap). MRI visual features, including diameter, calcification, shape (regular or irregular), margin (smooth or glitch), and strengthening (uniform or uneven), were independently evaluated by two radiologists with more than ten years of work experience. Tumor heterogeneity parameters acquisition and analysis Regions of interest (ROIs) were delineated manually on a single section containing the largest cross-section of tumor regions and tumor enhancement kinetics were retrospectively assessed by using the B.K. software (GE healthcare) automatically. The kinetic parameters included max peak enhancement, peak enhancement ratio, persistent, plateau, and washout. The max peak enhancement is the maximum value of the difference map before and after enhancement. The peak enhancement ratio is the ratio of the difference map before and after enhancement to the flat scan image. To investigate intra-tumoral heterogeneity within the tumor, the delayed enhancement kinetics automatically measured with the B.K. software was used to quantify the kinetic heterogeneity (a measure of heterogeneity in the proportions of tumor pixels with delayed washout, plateau, and persistent components in a tumor). Radiomics features extraction Original images were exported from the MRI system workstation in digital imaging and communications in medicine (DICOM) format and imported into A.K. software (GE healthcare). The polygon tool was used to sketch along the tumor edge to generate ROIs. Histogram-based features, second-order texture features, gray level co-occurrence matrix (GLCM)-related texture features, gray level size zone matrix (GLSZM) features, and morphological features were automatically extracted. The least absolute shrinkage and selection operator (LASSO) logistic regression method using 10-fold cross-validation was applied to select the most useful predictive ALNM status-related features from the training cohort. A radiomics score was generated per patient using a linear combination of the chosen features weighted by the LASSO algorithm. Construction of nomogram and risk stratification The candidate variables with p < 0.05 in the univariate analysis were input into the multivariate binary logistic backward stepwise regression analysis to select the independent predictors. A nomogram was constructed based on the results. According to the score of the nomogram, the model was risk stratified using the unsupervised clustering K-Mean algorithm into high risk, medium risk, and low risk groups. Statistical Analysis The clinical factors, MRI visual characteristics, radiomics parameters, and kinetic features obtained from preoperative MRI data were compared on the basis of ALNM status. Categorical variables were compared by using the 2 test or Fisher exact test. For continuous variables, the Shapiro-Wilk test for normality and the Levene F test for equal variance were performed. If the data were normally distributed and exhibited equal variance, the Student t test was used. Otherwise, the Mann-Whitney U test was performed. Logistic regression analysis was used to reveal associations between kinetic features and ALNM. Variables with p < 0.05 at univariate analysis were included in the multivariate logistic regression analysis using the backward stepwise selection method. Construct a predictive model based on the selected independent risk factors and present it in the form of nomogram. Odds ratio (OR) and 95% confidence interval (CI) were estimated by using the model. To dichotomize kinetic features of ALND analysis, the optimal cutoff values were determined by maximizing the sum of sensitivity and specificity with receiver operating characteristic (ROC) curve analysis. To avoid overfitting, we used the 10-fold cross-validation method to determine cutoff values in the ROC curve. That is, the dataset was randomly divided into 10 equal parts. Nine subsets were used for training, and the remaining subset was used for testing. This process was repeated 10 times so that each subset was used once for testing. All statistical analyses were performed with R statistical software (version 3.3.2; R Foundation for Statistical Computing, Vienna, Austria). A p < 0.05 indicated statistical significance in all analyses. Results Patient demographics A total of 373 consecutive female patients with cN0 breast cancer from 2016 to 2020 in Hospital #blinded# were included in the study. All the patients divided into a training group and a test group in a ratio of 8:2. In the training cohort, 220 patients (49.2 9.29) were ALNM negative, 78 patients (46.7 9.95) were ALNM positive. In the test cohort, 56 patients (48.7 8.74) were ALNM negative, 19 patients (44.4 7.46) were ALNM positive. The diameter of the ALNM (+) group was significantly larger than that of the ALNM (-) group (2.07cm 1.11 vs 1.79cm 0.80, p = 0.043) in the training cohort. The probability of ALNM was higher when the margin was glitch ( p < 0.05) in the training and test cohorts. Table 1 lists the baseline characteristics of the study population based on ALNM status in the training and test cohorts. Kinetic Features according to ALNM Status The ratio value was significantly higher in the ALNM (+) group than in the ALNM (-) group ( p < 0.05) in the training and test cohorts. The mean value of the washout was higher in the ALNM (+) group than in the ALNM (-) group. In regard to max peak enhancement, persistent, and plateau, no differences were observed between the ALNM (-) and ALNM (+) groups (Table 1 ). Table 1 The baseline characteristics, MRI visual features, kinetic features of the breast cancer in the training and test cohorts. Training cohort Test cohort ALNM (-) n = 220 ALNM (+) n = 78 p ALNM (-) n = 56 ALNM (+) n = 19 p Age 49.2 ± 9.29 46.7 ± 9.95 0.049 48.7 ± 8.74 44.4 ± 7.46 0.047 Menopause 0.122 0.196 No 114 (51.8%) 49 (62.8%) 23 (41.1%) 4 (21.1%) Yes 106 (48.2%) 29 (37.2%) 33 (58.9%) 15 (78.9%) Size 1.79 ± 0.80 2.07 ± 1.11 0.043 1.81 ± 0.65 2.05 ± 0.41 0.070 Location 0.243 0.907 Right 88 (40.0%) 38 (48.7%) 24 (42.9%) 8 (42.1%) Left 122 (55.5%) 39 (50.0%) 29 (51.8%) 11 (57.9%) Both 10 (4.55%) 1 (1.28%) 3 (5.36%) 0 Calcification 0.641 0.747 No 30 (13.6%) 13 (16.7%) 13 (23.2%) 3 (15.8%) Yes 190 (86.4%) 65 (83.3%) 43 (76.8%) 16 (84.2%) Multiple 0.292 0.271 No 172 (78.2%) 66 (84.6%) 46 (82.1%) 18 (94.7%) Yes 48 (21.8%) 12 (15.4%) 10 (17.9%) 1 (5.26%) Shape 0.070 1.000 Regular 124 (56.4%) 34 (43.6%) 20 (35.7%) 7 (36.8%) Irregular 96 (43.6%) 44 (56.4%) 36 (64.3%) 12 (63.2%) Margin < 0.001 0.001 Regular 137 (62.3%) 12 (15.4%) 32 (57.1%) 2 (10.5%) Irregular 83 (37.7%) 66 (84.6%) 24 (42.9%) 17 (89.5%) Strengthen 0.109 0.851 No 121 (55.0%) 34 (43.6%) 18 (32.1%) 5 (26.3%) Yes 99 (45.0%) 44 (56.4%) 38 (67.9%) 14 (73.7%) ER 0.010 0.344 Negative 43 (19.6%) 4(4.76%) 13 (16.5%) 2 (7.14%) Positive 177 (80.4%) 74 (95.2%) 66 (83.5%) 26 (92.9%) PR 0.039 0.369 Negative 44 (20.3%) 6 (7.94%) 16 (20.8%) 3 (10.7%) Positive 176 (79.7%) 72 (92.1%) 61 (79.2%) 25 (89.3%) Her 2 0.729 0.137 Negative 189 (85.9%) 65 (83.1%) 66 (78.6%) 27 (93.1%) Positive 31 (14.1%) 13 (16.9%) 18 (21.4%) 2 (6.90%) Ki67 0.162 0.709 Lower expression 51(23.3%) 11 (14.1%) 22 (26.5%) 6 (20.7%) High expression 169 (76.7%) 67 (85.9%) 61 (73.5%) 23 (79.3%) LVI < 0.001 < 0.001 Negative 59 (26.7%) 59 (75.4%) 11 (13.8%) 23 (79.3%) Positive 161 (73.3%) 19 (24.6%) 69 (86.2%) 6 (20.7%) MPE 226 ± 88.2 223 ± 88.9 0.829 215 ± 84.7 244 ± 111 0.296 Persistent 0.52 ± 0.27 0.47 ± 0.26 0.120 0.54 ± 0.28 0.48 ± 0.28 0.436 Plateau 0.09 ± 0.09 0.08 ± 0.12 0.771 0.09 ± 0.08 0.08 ± 0.06 0.869 Ratio 11.1 ± 14.0 17.2 ± 23.0 0.030 4.61 ± 4.36 15.1 ± 17.1 0.017 Washout 0.39 ± 0.30 0.52 ± 0.27 0.001 0.31 ± 0.30 0.44 ± 0.33 0.171 Rad-Score 0.26 ± 0.02 0.28 ± 0.05 0.002 0.26 ± 0.03 0.27 ± 0.03 0.387 MRI, magnetic resonance imaging; ALNM, axillary lymph node metastasis; ER, estrogen receptor; PR, progesterone receptor; Her 2, human epidermal growth factor receptor 2; LVI, lymphovascular invasion; MPE, max peak enhancement. Radiomics feature screening and radiomics score In the training cohort, 206 features were extracted from the original DCE-MRI images, including 42 histogram-based features, 334 second-order texture features, 144 GLCM-related texture features, 11 GLSZM features, and 9 morphological features. Two potential features were chosen in the training cohort with nonzero coefficients in the 10-fold cross-validation LASSO logistic regression model. These 2 features were used to calculate the radiomics score. The radiomics scores of ALNM (+) were 0.28 ± 0.05 and 0.27 ± 0.03 in the training and test cohorts and 0.26 ± 0.02 and 0.26 ± 0.03 for ALNM (-) patients in the training and test cohorts, respectively (Table 1 ). Prediction model development Univariate analysis was performed for each variable in the training cohort. Age, size, persistent, plateau, ratio, washout, and radiomics score were statistically associated with ALNM ( p ranged from 0.041 to 0.000) (Table 2 ). Table 2 Multivariate Logistic regression analysis of training cohort. Parameters Univariate analysis Multivariate analysis OR 95% CI p OR 95% CI p Age 2.028 1.202–3.421 0.008 1.925 1.010–3.670 0.047 Size 2.01 1.099–3.678 0.023 1.319 0.631–2.756 0.462 Location 0.676 0.421–1.084 0.104 Calcification 0.789 0.388–1.604 0.514 Multiple 0.652 0.326–1.303 0.226 Shape 1.672 0.993–2.814 0.053 Margin 9.078 4.633–17.791 0.000 12.199 5.745–25.902 0.000 Strengthen 1.582 0.940–2.662 0.084 MPE 1.312 0.767–2.243 0.322 Persistent 2.432 1.259-4.700 0.008 0.865 0.281–2.659 0.8 Plateau 0.557 0.318–0.976 0.041 0.713 0.310–1.639 0.425 Washout 3.735 1.820–7.666 0.000 5.309 2.289–12.312 0.000 Ratio 3.527 1.650–7.538 0.001 6.264 2.256–17.391 0.000 Rad-Score 3.027 1.719–5.332 0.000 3.735 1.888–7.391 0.000 OR, Odds ratio; CI, confidence interval; MPE, max peak enhancement. Furthermore, a multivariate binary logistic regression analysis identified that age (OR, 1.92526; 95% CI, 1.01001–3.6699; p , 0.047), margin (OR, 12.19893; 95% CI, 5.74553–25.9008; p < 0.001), ratio (OR, 6.26351; 95% CI, 2.25591–17.3906; p < 0.001), washout (OR, 5.30904 ; 95% CI, 2.28941–12.3115; p < 0.001), and radiomics scores (OR 3.73544; 95% CI, 1.88801–7.3906; p < 0.001) were independent risk predictors of ALNM (Table 2 ). The Hosmer-Lemeshow test showed that the p -value was 0.4569, indicating that the model had increased the goodness of fit. The variance inflation factor of each predictor was less than 10, and the corresponding tolerance was more than 0.1; therefore, there was no multicollinearity among these predictors (Table 3 ). Table 3 Multicollinearity assessment in the prediction model based on the independent predictors. Predictors Collinearity Statistics Tolerance VIF Age 0.719 1.03 Margin 0.922 1.08 Ratio 0.927 1.08 Washout 0.936 1.07 Rad-Score 0.946 1.06 VIF, variance inflation factor. The cut-off value of each DCE-MRI quantitative parameter was displayed in Table 4 , and these parameters were converted from continuous variables to categorical variables accordingly. Table 4 Prediction of each quantitative parameter for ALNM in patients with breast cancer. Parameters Cut-off value AUC (95% CI) p Age > 46 0.589 (0.531–0.646) 0.0189 Diameter > 1.4 0.567 (0.509–0.624) 0.0645 MPE > 178 0.509 (0.451–0.567) 0.8134 Persistent > 0.7816 0.549 (0.491–0.607) 0.1855 Plateau > 0.07362 0.538 (0.479–0.595) 0.3056 Ratio > 38 0.543 (0.484–0.600) 0.2786 Washout > 0.18 0.610 (0.552–0.665) 0.0026 Rad-Score > 0.2506 0.652 (0.595–0.706) < 0.0001 AUC, area under the curve; CI, confidence interval; MPE, max peak enhancement. A nomogram was produced by incorporating the above five independent predictors (Fig. 2 ). It showed good discrimination with an area under the curve (AUC) of 0.857 (95% CI, 0.812–0.894) (Table 5 and Fig. 3 A). The calibration curve showed good agreement between the nomogram-estimated probability of ALNM and the actual ALNM rate in the training cohort, with a mean absolute error of 0.021 (Fig. 3 D and Table 6 ). Table 5 The model performance in estimating the risk of ALNM in patients with breast cancer. Parameters Training cohort Test cohort Cut-off value 0.3 N/A AUC 0.857 (0.812–0.894) 0.858 (0.758–0.928) Sensitivity (%) 78.21 (67.4–86.8) 94.74 (74.0–99.9) Specificity (%) 82.27 (76.6–87.1) 71.43 (57.8–82.7) Accuracy (%) 81.21 77.33 PPV (%) 61.0 (53.5–68.0) 52.9 (42.3–63.3) NPV (%) 91.4 (87.4–94.2) 97.6 (85.5–99.6) PLR (%) 4.41 (3.2-6.0) 3.32 (2.2–5.1) NLR (%) 0.26 (0.2–0.4) 0.074 (0.01–0.5) Note: Data in parentheses is 95% CI. ALNM, axillary lymph node metastasis; AUC, area under the curve; PPV, positive predictive value; NPV, negative predictive value; PLR, positive likelihood ratio; NLR, negative likelihood ratio; CI, confidence interval. Table 6 The calibration efficiency of the prediction nomogram in the training and test cohorts. MAE MSE 0.9 QoAB Training cohort (n = 298) 0.021 6e-04 0.037 Test cohort (n = 75) 0.056 0.00767 0.127 MAE, mean absolute error; MSE, mean squared error; QoAB, quantile of absolute error. In addition, in the training cohort, the cutoff value of 0.3 was selected to distinguish the presence of ALNM, with a sensitivity of 78.21%, specificity of 82.37%, accuracy of 81.21%, positive predictive value (PPV) of 61.0%, negative predictive value (NPV) of 91.4%, positive likelihood ratio (PLR) of 4.41, and negative likelihood ratio (NLR) of 0.26 (Table 5 ). Risk stratification system of the nomogram In the training cohort, the cut-off value of 0.3 was selected to distinguish the presence of ALNM (Fig. 4 ). The unsupervised clustering K-Mean algorithm was used to risk stratify the model based on the scores from the nomogram, specifically: 0–78.874 points for the low-risk group, 92.929–166.74 points for the medium-risk group, and 178.874–318.965 points for the high-risk group. Prediction model validation Good discrimination with an AUC of 0.858 (95% CI, 0.758–0.928) was achieved in the test cohort (Table 5 and Fig. 3 B). Good calibration was also confirmed, with a mean absolute error of 0.056 (Fig. 3 D and Table 6 ). The decision curve revealed that if the threshold probability of a patient or physician was more significant than 7%, more advantages would be added by using the nomogram to estimate ALNM in breast cancer patients (Fig. 5 ). An example of predicting ALNM (-) correctly and one predicting ALNM(+) correctly were presented, respectively, to illustrate the clinical utility of the constructed nomogram (Fig. 6 and Fig. 7 ). For clinical use, age was determined by drawing a line straight up to the point axis to establish the score associated with the age. Next, this process was repeated for the other four covariates (washout, radiomics score, ratio, and margin). The scores of each covariate were added, and the total score was located on the total score points axis. Last, a line was drawn straight down to the risk of the ALNM axis to obtain the probability. Each parameter had corresponding value (points) that appear in upper toolbar as following: “Age 46 years old” = 26 points, “Washout > 0.18” = 67 points, “Radiomics score > 0.2506” = 53 points, “Ratio > 38” = 74 points, and “Margin was with glitch” = 100 points. A summarized total was applied on the bottom scale to obtain the probability of ALNM. Any probability greater than 0.3 (about 176 points) was compatible with ALNM. ALNM, axillary lymph node metastasis. Discussion In the current study, we investigated the feasibility and accuracy of the radiomics-based prediction model for prediction of ALNM in patients with breast cancer based on DCE-MRI images of the primary tumor. Our study has three significant findings. First, the radiomics nomogram, based on age, margin, ratio, washout, and radiomics score, showed a favorable ability to discriminate between ALNM (+) and ALNM (-), with AUC values of 0.857 and 0.858 in the training and test cohorts, respectively. Second, a higher degree of kinetic heterogeneity was associated with ALNM. The multivariate analysis showed that higher values of kinetic heterogeneity (ratio [OR 6.26351] and washout [OR 5.30904]), as determined with the nomogram, were associated with ALNM in women with breast cancer. Third, in the stratification of ALNM risk in women with breast cancer and thus allow therapies to be tailored based on individual risk level. The omission of axillary lymph node dissection might be justified in this subgroup of women with low-risk group. Breast cancer is a heterogeneous tumor with intra-tumoral temporal and spatial variation in cellularity, angiogenesis, and extravascular extracellular matrix [ 5 ]. It's a truism that the intra-tumor heterogeneity is associated with metastasis and poor prognosis due to inherently invasive biological behavior. However, it remains a challenge to quantify intra-tumoral heterogeneity in a noninvasive way before surgery. Tumor angiogenesis is one of the prerequisites for tumor progression and metastasis, and affects the uptake of contrast media within a tumor during DCE MRI. Therefore, it seems plausible that there are associations between ALNM and the breast cancer kinetic features extracted from DCE-MRI. In the current study, we hypothesized that intra-tumoral heterogeneity might be reflected in tumor enhancement kinetics from DCE-MRI and the concrete characteristic could be quantified with B.K. software. We presumed that higher tumor kinetic heterogeneity as observed on DCE-MRI images might reflect highly heterogeneous tumors with temporal and spatial variation in angiogenesis and various histopathologic components, further leading to the occurrence of ALNM. Vascular endothelial growth factor (VEGF), as we all known, is a major stimulator of angiogenesis, which is frequently overexpressed in breast cancer [ 13 ]. Thus, the various histopathological components in angiogenesis may be reflected in the tumor enhancement kinetics features, which explains the higher kinetic heterogeneity observed in the present study. The new capillaries formed via angiogenesis are typically immature and more permeable than the normal vasculature, which may cause ALNM[ 14 ]. This could explain the higher kinetic heterogeneity value for the possibility of ALNM. Most notably, the five DCE-MRI-based kinetic heterogeneity collected in this study showed independent predictive value in predicting ALNM, suggesting that ratio and washout could potentially identify the characteristics of intra-tumor heterogeneity driving aggressive tumor behavior. Patients assigned to the ALNM high-risk group showed increased heterogeneity, corroborating the hypothesis that tumor heterogeneity is related to aggressive tumor behavior. On the other hand, max peak enhancement, persistent, and plateau were not retained in the final model, which suggesting that DCE-MRI-based intra-tumor heterogeneity does not equal pathologic intra-tumor heterogeneity but can be used in prediction models as a valuable parameter. The radiomics score was another independently risk factor for predicting ALNM. Establishing a radiomics score with LASSO has demonstrated excellent results in predicting lymph node metastasis in papillary thyroid carcinoma [ 15 ], cervical cancer [ 16 ], pancreatic carcinoma [ 17 ], rectal cancer [ 18 ], and lung cancer [ 19 ]. Radiomics characteristics are closely related to the microstructure and biological behavior of the tumors [ 16 ]. The radiomics score is based on the high-dimensional and statistical features, which were extracted from primary breast tumors. In this study, two radiomics features were used to calculate the radiomics score. These features represent the texture information of tumors, which is highly associated with tumor heterogeneity [ 20 ]. Age and spicule also showed a significant relationship with ALNM, which is consistent with previous research [ 7 ].The rapid tumor growth rate exceeds the nutritional capacity of its blood supply, which is a possible reason of tumor glitch. A glitch is an invasion of the cell edge caused by factors external to the cell, which is often associated with rapid tumor growth and metastasis of breast cancer. More noteworthy was the risk stratification of the nomogram, which increased its clinical utility. For patients with low risk, we did not recommend to perform axillary lymph node dissection, but regular ultrasound follow-up, in order to reduce the surgical wound area and reduce the incidence of postoperative complications. For high-risk patients, we suggested that sentinel lymph node biopsy was not necessary and axillary lymph node dissection could be performed directly, thus reducing the intraoperative waiting time of patients and providing sufficient time for doctors to formulate reasonable preoperative surgical methods. Our study represents a preliminary test of our underlying hypothesis and has several limitations including a retrospective design. First, this study included a small number of patients from a single institution. Multi-institutional prospective studies would help to verify the results. Additional studies with longer follow-up and a larger population are needed to confirm the clinical utility of the nomogram. Second, the extraction of primary tumor features was semi-automatic segmentation, whether it was DCE-MRI or radiomics, which had errors to a certain extent. In the future, more prepared image acquisitions could be performed with fully automatic segmentation methods that have already been applied[ 21 , 22 ]. Third, tumor hypoxia can be estimated by hypoxic imaging techniques. The correlation between tumor hypoxia and molecular markers is a potential and interesting research on ALNM. Finally, integrating DCE-MRI-based heterogeneity with tumor histopathology and molecular subtypes holds the promise for better ALNM prediction and improved clinical decision making for individual patients. Conclusions In conclusion, we have proposed a radiomics-based nomogram for predicting ALNM in patients with breast cancer. Moreover, risk stratification can assist in the formulation of individualized treatment plans for patients in clinical practice. The quantitative assessment of DCE-MRI-based heterogeneity may provide biological information related to the aggressiveness of breast cancer that are important determinants of ALNM, and may potentially allow for more individualized treatment strategies. Furthermore, the advantages of the DCE-MRI-based heterogeneity assessment in this study are its noninvasive nature and the potential for easy application in routine clinical practice. Abbreviations AUC Area under the curve ALN Axillary lymph node ALND Axillary lymph node dissection ALNM Axillary lymph node metastasis CI Confidence interval DCE-MRI Dynamic contrast-enhanced magnetic resonance imaging DICOM Digital imaging and communications in medicine ER Estrogen receptor GLCM Gray level co-occurrence matrix GLSZM Gray level size zone matrix HER 2 Human epidermal growth factor receptor 2 LASSO Least absolute shrinkage and selection operator LVI Lymphovascular invasion MAE Mean absolute error MPE Max peak enhancement MRI Magnetic resonance imaging MSE Mean squared error NLR Negative likelihood ratio NPV Negative predictive value OR Odds ratio PLR Positive likelihood ratio PPV Positive predictive value PR Progesterone receptor QoAB Quantile of absolute error ROC Receiver operating characteristic ROI Regions of interest VEGF Vascular endothelial growth factor VIF Variance inflation factor Declarations Ethics approval and consent to participate This study was approved by the institutional review board of Jilin Cancer Hospital (IRB No.2023-001-01) and informed consent was waivered due to retrospective nature. Consent for publication Not applicable. Availability of data and materials All authors have reviewed the final version of the paper and would like to take public responsibility for its content. Competing interests The authors declare that they have no competing interests. Funding This study was supported by Science and Technology Developing Plan of Jilin Province (No.20230203093SF) and Health Technology Capability Enhancement Project of Jilin Province (No.2022JC022). Authors' contributions SS conceived and designed this study. TS and DY carried out the collection and assembly of data and drafted the manuscript. MY, JY and HZ were responsible for feature extraction and statistical work. All authors reviewed the manuscript. Acknowledgements Not applicable. Authors' information Department of Radiology, Jilin Cancer Hospital, Changchun, Jilin, China Tongxu Shen, Dingli Ye, Ming Yao, Jieqiong Yan, Han Zhang & Shuangyan Sun References Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer Statistics. Cancer J Clin. 2021;71(1):7–33. Samiei S, Granzier RWY, Ibrahim A et al. Dedicated Axillary MRI-Based Radiomics Analysis for the Prediction of Axillary Lymph Node Metastasis in Breast Cancer. Cancers. 2021; 13(4). Zhang X, Yang Z, Cui W, et al. Preoperative prediction of axillary sentinel lymph node burden with multiparametric MRI-based radiomics nomogram in early-stage breast cancer. Eur Radiol. 2021;31(8):5924–39. Han L, Zhu Y, Liu Z, et al. Radiomic nomogram for prediction of axillary lymph node metastasis in breast cancer. Eur Radiol. 2019;29(7):3820–9. Kim JY, Kim JJ, Hwangbo L, et al. Kinetic Heterogeneity of Breast Cancer Determined Using Computer-aided Diagnosis of Preoperative MRI Scans: Relationship to Distant Metastasis-Free Survival. Radiology. 2020;295(3):517–26. Fan M, Zhang P, Wang Y, et al. Radiomic analysis of imaging heterogeneity in tumours and the surrounding parenchyma based on unsupervised decomposition of DCE-MRI for predicting molecular subtypes of breast cancer. Eur Radiol. 2019;29(8):4456–67. Kim JJ, Kim JY, Suh HB, et al. Characterization of breast cancer subtypes based on quantitative assessment of intratumoral heterogeneity using dynamic contrast-enhanced and diffusion-weighted magnetic resonance imaging. Eur Radiol. 2021. 10.1007/s00330-021-08166-4 . Zhao R, Ma WJ, Tang J, et al. Heterogeneity of enhancement kinetics in dynamic contrast-enhanced MRI and implication of distant metastasis in invasive breast cancer. Clin Radiol. 2020;75(12):961. e925-961 e932. Chitalia RD, Rowland J, McDonald ES, et al. Imaging Phenotypes of Breast Cancer Heterogeneity in Preoperative Breast Dynamic Contrast Enhanced Magnetic Resonance Imaging (DCE-MRI) Scans Predict 10-Year Recurrence. Clin cancer research: official J Am Association Cancer Res. 2020;26(4):862–9. Collins GS, Reitsma JB, Altman DG, Moons KGM, members of the Tg. Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD): The TRIPOD Statement. Eur Urol. 2015;67(6):1142–51. Bitencourt AGV, Gibbs P, Rossi Saccarelli C, et al. MRI-based machine learning radiomics can predict HER2 expression level and pathologic response after neoadjuvant therapy in HER2 overexpressing breast cancer. EBioMedicine. 2020;61:103042. Lee JY, Lee KS, Seo BK et al. Radiomic machine learning for predicting prognostic biomarkers and molecular subtypes of breast cancer using tumor heterogeneity and angiogenesis properties on MRI. Eur Radiol. 2021; 1–11. Malekan M, Ebrahimzadeh MA. Vascular Endothelial Growth Factor Receptors [VEGFR] as Target in Breast Cancer Treatment: Current Status in Preclinical and Clinical Studies and Future Directions. Curr Top Med Chem. 2022;22(11):891–920. Siemann DW, Chaplin DJ, Horsman MR. Realizing the Potential of Vascular Targeted Therapy: The Rationale for Combining Vascular Disrupting Agents and Anti-Angiogenic Agents to Treat Cancer. Cancer Invest. 2017; 14;35(8):519–534. Jiang M, Li C, Tang S, et al. Nomogram Based on Shear-Wave Elastography Radiomics Can Improve Preoperative Cervical Lymph Node Staging for Papillary Thyroid Carcinoma. Thyroid: official J Am Thyroid Association. 2020;30(6):885–97. Jin X, Ai Y, Zhang J, et al. Noninvasive prediction of lymph node status for patients with early-stage cervical cancer based on radiomics features from ultrasound images. Eur Radiol. 2020;30(7):4117–24. Bian Y, Guo S, Jiang H, et al. Relationship Between Radiomics and Risk of Lymph Node Metastasis in Pancreatic Ductal Adenocarcinoma. Pancreas. 2019;48(9):1195–203. Nakanishi R, Akiyoshi T, Toda S, et al. Radiomics Approach Outperforms Diameter Criteria for Predicting Pathological Lateral Lymph Node Metastasis After Neoadjuvant (Chemo)Radiotherapy in Advanced Low Rectal Cancer. Ann Surg Oncol. 2020;27(11):4273–83. Xie Y, Zhao H, Guo Y, et al. A PET/CT nomogram incorporating SUVmax and CT radiomics for preoperative nodal staging in non-small cell lung cancer. Eur Radiol. 2021;31(8):6030–8. Hu HT, Wang Z, Huang XW, et al. Ultrasound-based radiomics score: a potential biomarker for the prediction of microvascular invasion in hepatocellular carcinoma. Eur Radiol. 2019;29(6):2890–901. Jiao H, Jiang X, Pang Z et al. Deep Convolutional Neural Networks-Based Automatic Breast Segmentation and Mass Detection in DCE-MRI. Comput Math Methods Med. 2020; 2020:2413706. Rahimpour M, Saint Martin MJ, Frouin F, et al. Visual ensemble selection of deep convolutional neural networks for 3D segmentation of breast tumors on dynamic contrast enhanced MRI. Eur Radiol. 2023;33(2):959–69. Additional Declarations No competing interests reported. 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-4445164","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":307401659,"identity":"21c2c02d-fddf-441c-b945-93fcda9e703e","order_by":0,"name":"Tongxu Shen","email":"","orcid":"","institution":"Jilin Cancer Hospital","correspondingAuthor":false,"prefix":"","firstName":"Tongxu","middleName":"","lastName":"Shen","suffix":""},{"id":307401660,"identity":"a7d885d7-75a3-4966-91fa-1953fd1589b4","order_by":1,"name":"Dingli Ye","email":"","orcid":"","institution":"Jilin Cancer Hospital","correspondingAuthor":false,"prefix":"","firstName":"Dingli","middleName":"","lastName":"Ye","suffix":""},{"id":307401661,"identity":"65d8f205-a569-4927-b3a4-d3edf9e5d750","order_by":2,"name":"Ming Yao","email":"","orcid":"","institution":"Jilin Cancer Hospital","correspondingAuthor":false,"prefix":"","firstName":"Ming","middleName":"","lastName":"Yao","suffix":""},{"id":307401662,"identity":"67711b36-06ce-4353-bce7-4323df41bd1f","order_by":3,"name":"Jieqiong Yan","email":"","orcid":"","institution":"Jilin Cancer Hospital","correspondingAuthor":false,"prefix":"","firstName":"Jieqiong","middleName":"","lastName":"Yan","suffix":""},{"id":307401663,"identity":"93de599f-aec3-45b3-b558-1c787c2be1d3","order_by":4,"name":"Han Zhang","email":"","orcid":"","institution":"Jilin Cancer Hospital","correspondingAuthor":false,"prefix":"","firstName":"Han","middleName":"","lastName":"Zhang","suffix":""},{"id":307401664,"identity":"75913060-3140-4177-a6e9-bbb891e6246d","order_by":5,"name":"Shuangyan Sun","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/UlEQVRIiWNgGAWjYJACZjgrweC/HBt7+wEStHyoYDbm4zmTQLwWxhlnmBPnSTgY4FUuPyP38OfCNrs8+YjkZ49529jS2yQYEhh+VGzDqcXgRl6a9My25GLDG2nmxrxtPLlt0o0HGHvO3MatRSLHjJm3jTlx44wEM2neNoncNpkDCcyMbbi1yM/IMf7M21YP1JL+DajFIJ1NIsEArxaGGzkGQJWHE+cDrZOccSYhgaAWgzNvzKR5zh1P3MDzpkziQ8UBwzZgIB/E5xf5dqDDeMqqE+e3p28Dmn9AXr69/eCDHxV4HAa37kICgnOAsHqQdf3EqRsFo2AUjIIRCAD8TldSkaeLjAAAAABJRU5ErkJggg==","orcid":"","institution":"Jilin Cancer Hospital","correspondingAuthor":true,"prefix":"","firstName":"Shuangyan","middleName":"","lastName":"Sun","suffix":""}],"badges":[],"createdAt":"2024-05-19 16:57:46","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4445164/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4445164/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":57722133,"identity":"e33cdbcb-b28a-4e73-a666-20f04298f9f3","added_by":"auto","created_at":"2024-06-04 19:01:15","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":892184,"visible":true,"origin":"","legend":"\u003cp\u003eThe flowchart of the patient inclusion and exclusion criteria in the current study.\u003c/p\u003e\n\u003cp\u003eDCE-MRI, dynamic contrast-enhanced magnetic resonance imaging; ALNM, axillary lymph node metastasis.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-4445164/v1/36425ad23e024e7fe3b0bf5a.png"},{"id":57722132,"identity":"ef4a6e4e-4d83-4864-9c6b-e4c77e1e2480","added_by":"auto","created_at":"2024-06-04 19:01:15","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":97980,"visible":true,"origin":"","legend":"\u003cp\u003eThe nomogram of the study.\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-4445164/v1/d993138abc0df7d206d8df5f.png"},{"id":57722543,"identity":"037a121f-b0a6-4c7d-9c50-246c13609839","added_by":"auto","created_at":"2024-06-04 19:09:15","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":348604,"visible":true,"origin":"","legend":"\u003cp\u003eThe ROC curves (A, B) and calibration curves (C, D) of the training and test cohorts.\u003c/p\u003e\n\u003cp\u003eROC, receiver operating characteristic.\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-4445164/v1/6155c2797ad83461b35e426f.png"},{"id":57722544,"identity":"bc2a6f0d-1e6a-4c04-bd13-566601024e15","added_by":"auto","created_at":"2024-06-04 19:09:15","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":179130,"visible":true,"origin":"","legend":"\u003cp\u003eThe cutoff value of the nomogram.\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-4445164/v1/f441287ec53d505d2542efdc.png"},{"id":57722139,"identity":"a285babf-9f8a-46ea-8f66-64be87eab670","added_by":"auto","created_at":"2024-06-04 19:01:15","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":93097,"visible":true,"origin":"","legend":"\u003cp\u003eDecision curve analysis of the nomogram in the current study.\u003c/p\u003e","description":"","filename":"Figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-4445164/v1/f20022ad4e44224c69c8b121.png"},{"id":57722136,"identity":"c0d54db6-ec04-4a0f-9d03-8b5fd747037a","added_by":"auto","created_at":"2024-06-04 19:01:15","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":392114,"visible":true,"origin":"","legend":"\u003cp\u003eOne example of correct prediction of ALNM+ in patients with breast cancer.\u003c/p\u003e\n\u003cp\u003eA-F, A 52-year-old female patient was admitted to the hospital for further treatment after physical examination found a nodule in the right breast. MRI visual features showed a nodule was with glitch (A). DCE-MRI kinetic heterogeneity parameters showed: washout was 0.7198 and ratio was 0 (B). The radiomics score was＞0.652 (C-F). A vertical line of each variable was drawn. The values on the “Points” scale intersected by the lines were added to obtain total points (0 + 67 + 53 + 74 + 100 = 294). The total points \u0026gt; 176 points, considered as a high-risk patient (G). The graph revealed that the risk of ALNM was nearly 90% by drawing a vertical line on the “Total points” scale. Postoperative pathological results showed that right breast non-invasive ductal carcinoma and metastatic carcinoma was found in the axillary.\u003c/p\u003e\n\u003cp\u003eALNM, axillary lymph node metastasis; DCE-MRI, dynamic contrast-enhanced magnetic resonance imaging.\u003c/p\u003e","description":"","filename":"Figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-4445164/v1/86c4ef22cfb60365c7cf8fae.png"},{"id":57722138,"identity":"941a2205-5f94-4dd3-ada4-3e890cdf27b0","added_by":"auto","created_at":"2024-06-04 19:01:15","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":353424,"visible":true,"origin":"","legend":"\u003cp\u003eOne example of correct prediction of ALNM- in patients with breast cancer.\u003c/p\u003e\n\u003cp\u003eA-F, A 41-year-old female patient was admitted to the hospital for further treatment after physical examination found a nodule in the left breast. MRI visual features showed a nodule was without glitch (A). DCE-MRI kinetic heterogeneity parameters showed: washout was 0 and ratio was 44 (B). The radiomics score was ＜0.652 (C-F). A vertical line of each variable was drawn. The values on the “Points” scale intersected by the lines were added to obtain total points (26 + 0 + 0 + 74 + 0 = 100). The total points \u0026lt; 176 points, considered as a low-risk patient (G). The graph revealed that the risk of ALNM was lower than 10% by drawing a vertical line on the “Total points” scale. Postoperative pathological results showed that right breast non-invasive ductal carcinoma without axillary lymph node metastasis.\u003c/p\u003e\n\u003cp\u003eALNM, axillary lymph node metastasis; DCE-MRI, dynamic contrast-enhanced magnetic resonance imaging.\u003c/p\u003e","description":"","filename":"Figure7.png","url":"https://assets-eu.researchsquare.com/files/rs-4445164/v1/180b011c9fd2b7e70a6f5884.png"},{"id":74156302,"identity":"639c5b8d-06de-4d13-ac17-1afebb802ec2","added_by":"auto","created_at":"2025-01-18 21:31:20","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3537439,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4445164/v1/644fc281-aa47-4e56-a7cf-ce61cf52bb17.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Kinetic heterogeneity is associated with axillary lymph node metastasis in cN0 breast cancer based on dynamic contrast-enhanced magnetic resonance imaging radiomics nomogram","fulltext":[{"header":"Introduction","content":"\u003cp\u003eBreast cancer accounts for 30% of newly diagnosed female cancers, ranking second in the deaths from cancers in women, second only to lung cancer in 2021 [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. The axillary lymph node (ALN) status affects the locoregional recurrence and overall survival rate [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Meanwhile, ALN status determines the scope of the surgery and the potential need for neoadjuvant chemotherapy. However, owing to the invasive nature of ALN dissection (ALND), the patients face considerable complications [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Therefore, preoperative noninvasive prediction of ALN metastasis (ALNM) is particularly important.\u003c/p\u003e \u003cp\u003ePhysical examination, mammography, ultrasound, and magnetic resonance imaging (MRI) are all commonly used for preoperative diagnosis of breast cancer; however, their ability to assess ALN is not satisfactory [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Because dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is independent of breast density, it is currently a sensitive and specific imaging technique for the diagnosis of breast cancer. It depicts the morphologic features of breast tumors and reveals dynamics of enhancement, which may reflect angiogenesis [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Breast cancer is highly heterogeneous in nature. The intra-tumoral heterogeneity is associated with molecular subtypes [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e], metastasis [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e], prognosis [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e] and recurrence [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e] in breast cancer, which may be due to intrinsic aggressive biology or treatment resistance.\u003c/p\u003e \u003cp\u003eRadiomics transforms medical images into higher dimensional data, which has been widely used in clinical medicine. Prediction models are becoming recognized as valuable tools in recent years and are recommended in clinical practice guidelines [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Radiomics-based prediction model has been applied in the prediction of sentinel lymph node burden [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e], human epidermal growth factor receptor 2 (HER 2) expression level [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e], prognostic biomarkers and molecular subtypes [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e] of breast cancer in the recent years. However, whether DCE-MRI radiomics-based nomogram could predict ALNM before surgery accurately, and whether it can be explained from the perspective of intra-tumoral heterogeneity, there are few relevant researches at present.\u003c/p\u003e \u003cp\u003eWe hypothesized that kinetic features of breast cancer and the heterogeneity of these features in preoperative DCE-MRI were associated with ALNM in women with breast cancer. The purposes of the study were, first, to extract the key elements for predicting ALNM, including clinical, DCE-MRI, and radiomics features. Second, to establish radiomics-based nomogram based on the key features for the prediction of ALNM. Third, to validate and evaluate the nomogram.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eThis retrospective study was approved by the institutional review board of Jilin Cancer Hospital (IRB No.2023-001-01). The requirement for informed consent was waive.\u003c/p\u003e\n\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n \u003ch2\u003eStudy population\u003c/h2\u003e\n \u003cp\u003eA retrospective review of medical records collected between 2016 and 2020 identified 2606 consecutive patients with breast cancer who underwent preoperative DCE-MRI. Patients were excluded if (a) male, (b) with incomplete clinical or pathological information, (c) with a history of other malignancy, (d) underwent excisional biopsy for diagnosis, (e) received neoadjuvant chemotherapy before surgery, (f) suspected of ALNM based on preoperative ultrasound or palpation, (g) with chemical shift artifacts or motion artifacts on the DCE-MRI images, (h) DCE-MRI examination time was over two weeks (Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e). For multiple lesions, the lesion with the largest diameter was selected for subsequent data analysis.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\n \u003ch2\u003eMRI acquisition and analysis\u003c/h2\u003e\n \u003cp\u003eMRI was performed in the prone position with a 3.0 T MRI system (Ingenia, Philips, Netherlands) using a dedicated eight-channel breast array coil, with the arms raised and the head advanced. Bilateral axial T2-weighted turbo spin-echo imaging and T1-weighted volumetric interpolated breath-hold examination with fat saturation were acquired. DCE-MRI images were acquired using a three-dimensional fat suppressed T1-weighted sequence and included one pre-contrast and five post-contrast images (repetition time, 4.6 ms; echo time, 2.0 ms; flip angle, 12\u0026deg;; field of view, 300\u0026times;380 mm\u003csup\u003e2\u003c/sup\u003e; matrix, 300\u0026times;380; section thickness, 1 mm; no gap).\u003c/p\u003e\n \u003cp\u003eMRI visual features, including diameter, calcification, shape (regular or irregular), margin (smooth or glitch), and strengthening (uniform or uneven), were independently evaluated by two radiologists with more than ten years of work experience.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\n \u003ch2\u003eTumor heterogeneity parameters acquisition and analysis\u003c/h2\u003e\n \u003cp\u003eRegions of interest (ROIs) were delineated manually on a single section containing the largest cross-section of tumor regions and tumor enhancement kinetics were retrospectively assessed by using the B.K. software (GE healthcare) automatically. The kinetic parameters included max peak enhancement, peak enhancement ratio, persistent, plateau, and washout. The max peak enhancement is the maximum value of the difference map before and after enhancement. The peak enhancement ratio is the ratio of the difference map before and after enhancement to the flat scan image. To investigate intra-tumoral heterogeneity within the tumor, the delayed enhancement kinetics automatically measured with the B.K. software was used to quantify the kinetic heterogeneity (a measure of heterogeneity in the proportions of tumor pixels with delayed washout, plateau, and persistent components in a tumor).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\n \u003ch2\u003eRadiomics features extraction\u003c/h2\u003e\n \u003cp\u003eOriginal images were exported from the MRI system workstation in digital imaging and communications in medicine (DICOM) format and imported into A.K. software (GE healthcare). The polygon tool was used to sketch along the tumor edge to generate ROIs. Histogram-based features, second-order texture features, gray level co-occurrence matrix (GLCM)-related texture features, gray level size zone matrix (GLSZM) features, and morphological features were automatically extracted. The least absolute shrinkage and selection operator (LASSO) logistic regression method using 10-fold cross-validation was applied to select the most useful predictive ALNM status-related features from the training cohort. A radiomics score was generated per patient using a linear combination of the chosen features weighted by the LASSO algorithm.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\n \u003ch2\u003eConstruction of nomogram and risk stratification\u003c/h2\u003e\n \u003cp\u003eThe candidate variables with \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05 in the univariate analysis were input into the multivariate binary logistic backward stepwise regression analysis to select the independent predictors. A nomogram was constructed based on the results. According to the score of the nomogram, the model was risk stratified using the unsupervised clustering K-Mean algorithm into high risk, medium risk, and low risk groups.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\n \u003ch2\u003eStatistical Analysis\u003c/h2\u003e\n \u003cp\u003eThe clinical factors, MRI visual characteristics, radiomics parameters, and kinetic features obtained from preoperative MRI data were compared on the basis of ALNM status. Categorical variables were compared by using the 2 test or Fisher exact test. For continuous variables, the Shapiro-Wilk test for normality and the Levene F test for equal variance were performed. If the data were normally distributed and exhibited equal variance, the Student \u003cem\u003et\u003c/em\u003e test was used. Otherwise, the Mann-Whitney U test was performed. Logistic regression analysis was used to reveal associations between kinetic features and ALNM. Variables with \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05 at univariate analysis were included in the multivariate logistic regression analysis using the backward stepwise selection method. Construct a predictive model based on the selected independent risk factors and present it in the form of nomogram. Odds ratio (OR) and 95% confidence interval (CI) were estimated by using the model.\u003c/p\u003e\n \u003cp\u003eTo dichotomize kinetic features of ALND analysis, the optimal cutoff values were determined by maximizing the sum of sensitivity and specificity with receiver operating characteristic (ROC) curve analysis.\u003c/p\u003e\n \u003cp\u003eTo avoid overfitting, we used the 10-fold cross-validation method to determine cutoff values in the ROC curve. That is, the dataset was randomly divided into 10 equal parts. Nine subsets were used for training, and the remaining subset was used for testing. This process was repeated 10 times so that each subset was used once for testing.\u003c/p\u003e\n \u003cp\u003eAll statistical analyses were performed with R statistical software (version 3.3.2; R Foundation for Statistical Computing, Vienna, Austria). A \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05 indicated statistical significance in all analyses.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\n \u003ch2\u003ePatient demographics\u003c/h2\u003e\n \u003cp\u003eA total of 373 consecutive female patients with cN0 breast cancer from 2016 to 2020 in Hospital #blinded# were included in the study. All the patients divided into a training group and a test group in a ratio of 8:2. In the training cohort, 220 patients (49.2 9.29) were ALNM negative, 78 patients (46.7 9.95) were ALNM positive. In the test cohort, 56 patients (48.7 8.74) were ALNM negative, 19 patients (44.4 7.46) were ALNM positive. The diameter of the ALNM (+) group was significantly larger than that of the ALNM (-) group (2.07cm 1.11 vs 1.79cm 0.80, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.043) in the training cohort. The probability of ALNM was higher when the margin was glitch (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05) in the training and test cohorts. Table \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e lists the baseline characteristics of the study population based on ALNM status in the training and test cohorts.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\n \u003ch2\u003eKinetic Features according to ALNM Status\u003c/h2\u003e\n \u003cp\u003eThe ratio value was significantly higher in the ALNM (+) group than in the ALNM (-) group (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05) in the training and test cohorts. The mean value of the washout was higher in the ALNM (+) group than in the ALNM (-) group. In regard to max peak enhancement, persistent, and plateau, no differences were observed between the ALNM (-) and ALNM (+) groups (Table \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003e\u003c/p\u003e\u0026nbsp;\u003ctable id=\"Tab1\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eThe baseline characteristics, MRI visual features, kinetic features of the breast cancer in the training and test cohorts.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\" colspan=\"4\"\u003e\n \u003cp\u003eTraining cohort\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colspan=\"3\"\u003e\n \u003cp\u003eTest cohort\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eALNM (-)\u003c/p\u003e\n \u003cp\u003en\u0026thinsp;=\u0026thinsp;220\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eALNM (+)\u003c/p\u003e\n \u003cp\u003en\u0026thinsp;=\u0026thinsp;78\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003eALNM (-)\u003c/p\u003e\n \u003cp\u003en\u0026thinsp;=\u0026thinsp;56\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eALNM (+)\u003c/p\u003e\n \u003cp\u003en\u0026thinsp;=\u0026thinsp;19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e49.2\u0026thinsp;\u0026plusmn;\u0026thinsp;9.29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e46.7\u0026thinsp;\u0026plusmn;\u0026thinsp;9.95\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003e0.049\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e48.7\u0026thinsp;\u0026plusmn;\u0026thinsp;8.74\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e44.4\u0026thinsp;\u0026plusmn;\u0026thinsp;7.46\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.047\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMenopause\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003e0.122\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.196\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e114 (51.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e49 (62.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e23 (41.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4 (21.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e106 (48.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e29 (37.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e33 (58.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e15 (78.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSize\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.79\u0026thinsp;\u0026plusmn;\u0026thinsp;0.80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2.07\u0026thinsp;\u0026plusmn;\u0026thinsp;1.11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003e0.043\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.81\u0026thinsp;\u0026plusmn;\u0026thinsp;0.65\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2.05\u0026thinsp;\u0026plusmn;\u0026thinsp;0.41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.070\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLocation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003e0.243\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.907\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRight\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e88 (40.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e38 (48.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e24 (42.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e8 (42.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLeft\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e122 (55.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e39 (50.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e29 (51.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e11 (57.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBoth\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e10 (4.55%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1 (1.28%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3 (5.36%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCalcification\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003e0.641\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.747\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e30 (13.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e13 (16.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e13 (23.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3 (15.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e190 (86.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e65 (83.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e43 (76.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e16 (84.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMultiple\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003e0.292\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.271\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e172 (78.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e66 (84.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e46 (82.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e18 (94.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e48 (21.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e12 (15.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e10 (17.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1 (5.26%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eShape\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003e0.070\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRegular\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e124 (56.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e34 (43.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e20 (35.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e7 (36.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIrregular\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e96 (43.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e44 (56.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e36 (64.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e12 (63.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMargin\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRegular\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e137 (62.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e12 (15.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e32 (57.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2 (10.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIrregular\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e83 (37.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e66 (84.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e24 (42.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e17 (89.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eStrengthen\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003e0.109\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.851\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e121 (55.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e34 (43.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e18 (32.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e5 (26.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e99 (45.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e44 (56.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e38 (67.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e14 (73.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eER\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003e0.010\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.344\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNegative\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e43 (19.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4(4.76%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e13 (16.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2 (7.14%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePositive\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e177 (80.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e74 (95.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e66 (83.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e26 (92.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003e0.039\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.369\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNegative\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e44 (20.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e6 (7.94%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e16 (20.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3 (10.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePositive\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e176 (79.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e72 (92.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e61 (79.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e25 (89.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eHer 2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003e0.729\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.137\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNegative\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e189 (85.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e65 (83.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e66 (78.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e27 (93.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePositive\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e31 (14.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e13 (16.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e18 (21.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2 (6.90%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eKi67\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003e0.162\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.709\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLower expression\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e51(23.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e11 (14.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e22 (26.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e6 (20.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eHigh expression\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e169 (76.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e67 (85.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e61 (73.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e23 (79.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLVI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNegative\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e59 (26.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e59 (75.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e11 (13.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e23 (79.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePositive\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e161 (73.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e19 (24.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e69 (86.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e6 (20.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMPE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e226\u0026thinsp;\u0026plusmn;\u0026thinsp;88.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e223\u0026thinsp;\u0026plusmn;\u0026thinsp;88.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003e0.829\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e215\u0026thinsp;\u0026plusmn;\u0026thinsp;84.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e244\u0026thinsp;\u0026plusmn;\u0026thinsp;111\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.296\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePersistent\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.52\u0026thinsp;\u0026plusmn;\u0026thinsp;0.27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.47\u0026thinsp;\u0026plusmn;\u0026thinsp;0.26\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003e0.120\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.54\u0026thinsp;\u0026plusmn;\u0026thinsp;0.28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.48\u0026thinsp;\u0026plusmn;\u0026thinsp;0.28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.436\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePlateau\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.09\u0026thinsp;\u0026plusmn;\u0026thinsp;0.09\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.08\u0026thinsp;\u0026plusmn;\u0026thinsp;0.12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003e0.771\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.09\u0026thinsp;\u0026plusmn;\u0026thinsp;0.08\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.08\u0026thinsp;\u0026plusmn;\u0026thinsp;0.06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.869\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRatio\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e11.1\u0026thinsp;\u0026plusmn;\u0026thinsp;14.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e17.2\u0026thinsp;\u0026plusmn;\u0026thinsp;23.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003e0.030\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4.61\u0026thinsp;\u0026plusmn;\u0026thinsp;4.36\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e15.1\u0026thinsp;\u0026plusmn;\u0026thinsp;17.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.017\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eWashout\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.39\u0026thinsp;\u0026plusmn;\u0026thinsp;0.30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.52\u0026thinsp;\u0026plusmn;\u0026thinsp;0.27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.31\u0026thinsp;\u0026plusmn;\u0026thinsp;0.30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.44\u0026thinsp;\u0026plusmn;\u0026thinsp;0.33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.171\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRad-Score\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.26\u0026thinsp;\u0026plusmn;\u0026thinsp;0.02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.28\u0026thinsp;\u0026plusmn;\u0026thinsp;0.05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003e0.002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.26\u0026thinsp;\u0026plusmn;\u0026thinsp;0.03\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.27\u0026thinsp;\u0026plusmn;\u0026thinsp;0.03\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.387\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003cp\u003e\u003c/p\u003e\n \u003cp\u003eMRI, magnetic resonance imaging; ALNM, axillary lymph node metastasis; ER, estrogen receptor; PR, progesterone receptor; Her 2, human epidermal growth factor receptor 2; LVI, lymphovascular invasion; MPE, max peak enhancement.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\n \u003ch2\u003eRadiomics feature screening and radiomics score\u003c/h2\u003e\n \u003cp\u003eIn the training cohort, 206 features were extracted from the original DCE-MRI images, including 42 histogram-based features, 334 second-order texture features, 144 GLCM-related texture features, 11 GLSZM features, and 9 morphological features. Two potential features were chosen in the training cohort with nonzero coefficients in the 10-fold cross-validation LASSO logistic regression model. These 2 features were used to calculate the radiomics score. The radiomics scores of ALNM (+) were 0.28\u0026thinsp;\u0026plusmn;\u0026thinsp;0.05 and 0.27\u0026thinsp;\u0026plusmn;\u0026thinsp;0.03 in the training and test cohorts and 0.26\u0026thinsp;\u0026plusmn;\u0026thinsp;0.02 and 0.26\u0026thinsp;\u0026plusmn;\u0026thinsp;0.03 for ALNM (-) patients in the training and test cohorts, respectively (Table \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\n \u003ch2\u003ePrediction model development\u003c/h2\u003e\n \u003cp\u003eUnivariate analysis was performed for each variable in the training cohort. Age, size, persistent, plateau, ratio, washout, and radiomics score were statistically associated with ALNM (\u003cem\u003ep\u003c/em\u003e ranged from 0.041 to 0.000) (Table \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003e\u003c/p\u003e\u0026nbsp;\u003ctable id=\"Tab2\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eMultivariate Logistic regression analysis of training cohort.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" rowspan=\"2\"\u003e\n \u003cp\u003eParameters\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colspan=\"3\"\u003e\n \u003cp\u003eUnivariate analysis\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" rowspan=\"2\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\" colspan=\"3\"\u003e\n \u003cp\u003eMultivariate analysis\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eOR\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e95% CI\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eOR\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e95% CI\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2.028\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.202\u0026ndash;3.421\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.008\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.925\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.010\u0026ndash;3.670\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.047\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSize\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2.01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.099\u0026ndash;3.678\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.023\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.319\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.631\u0026ndash;2.756\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.462\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLocation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.676\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.421\u0026ndash;1.084\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.104\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCalcification\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.789\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.388\u0026ndash;1.604\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.514\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMultiple\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.652\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.326\u0026ndash;1.303\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.226\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eShape\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.672\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.993\u0026ndash;2.814\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.053\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMargin\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e9.078\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4.633\u0026ndash;17.791\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e12.199\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e5.745\u0026ndash;25.902\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eStrengthen\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.582\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.940\u0026ndash;2.662\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.084\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMPE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.312\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.767\u0026ndash;2.243\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.322\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePersistent\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2.432\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.259-4.700\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.008\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.865\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.281\u0026ndash;2.659\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.8\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePlateau\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.557\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.318\u0026ndash;0.976\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.041\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.713\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.310\u0026ndash;1.639\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.425\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eWashout\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3.735\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.820\u0026ndash;7.666\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e5.309\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2.289\u0026ndash;12.312\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRatio\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3.527\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.650\u0026ndash;7.538\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e6.264\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2.256\u0026ndash;17.391\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRad-Score\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3.027\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.719\u0026ndash;5.332\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3.735\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.888\u0026ndash;7.391\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003cp\u003e\u003c/p\u003e\n \u003cp\u003eOR, Odds ratio; CI, confidence interval; MPE, max peak enhancement.\u003c/p\u003e\n \u003cp\u003eFurthermore, a multivariate binary logistic regression analysis identified that age (OR, 1.92526; 95% CI, 1.01001\u0026ndash;3.6699; \u003cem\u003ep\u003c/em\u003e, 0.047), margin (OR, 12.19893; 95% CI, 5.74553\u0026ndash;25.9008; \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001), ratio (OR, 6.26351; 95% CI, 2.25591\u0026ndash;17.3906; \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001), washout (OR, 5.30904 ; 95% CI, 2.28941\u0026ndash;12.3115; \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001), and radiomics scores (OR 3.73544; 95% CI, 1.88801\u0026ndash;7.3906; \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001) were independent risk predictors of ALNM (Table \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e). The Hosmer-Lemeshow test showed that the \u003cem\u003ep\u003c/em\u003e-value was 0.4569, indicating that the model had increased the goodness of fit. The variance inflation factor of each predictor was less than 10, and the corresponding tolerance was more than 0.1; therefore, there was no multicollinearity among these predictors (Table \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003e\u003c/p\u003e\u0026nbsp;\u003ctable id=\"Tab3\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eMulticollinearity assessment in the prediction model based on the independent predictors.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" rowspan=\"2\"\u003e\n \u003cp\u003ePredictors\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colspan=\"2\"\u003e\n \u003cp\u003eCollinearity Statistics\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eTolerance\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eVIF\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.719\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.03\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMargin\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.922\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.08\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRatio\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.927\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.08\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eWashout\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.936\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.07\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRad-Score\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.946\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.06\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003cp\u003e\u003c/p\u003e\n \u003cp\u003eVIF, variance inflation factor.\u003c/p\u003e\n \u003cp\u003eThe cut-off value of each DCE-MRI quantitative parameter was displayed in Table \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e, and these parameters were converted from continuous variables to categorical variables accordingly.\u003c/p\u003e\n \u003cp\u003e\u003c/p\u003e\u0026nbsp;\u003ctable id=\"Tab4\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003ePrediction of each quantitative parameter for ALNM in patients with breast cancer.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eParameters\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eCut-off value\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eAUC (95% CI)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026gt;\u0026thinsp;46\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.589 (0.531\u0026ndash;0.646)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.0189\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDiameter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026gt; 1.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.567 (0.509\u0026ndash;0.624)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.0645\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMPE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026gt; 178\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.509 (0.451\u0026ndash;0.567)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.8134\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePersistent\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026gt;\u0026thinsp;0.7816\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.549 (0.491\u0026ndash;0.607)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.1855\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePlateau\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026gt;\u0026thinsp;0.07362\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.538 (0.479\u0026ndash;0.595)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.3056\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRatio\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026gt; 38\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.543 (0.484\u0026ndash;0.600)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.2786\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eWashout\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026gt; 0.18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.610 (0.552\u0026ndash;0.665)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.0026\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRad-Score\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026gt;\u0026thinsp;0.2506\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.652 (0.595\u0026ndash;0.706)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003cp\u003e\u003c/p\u003e\n \u003cp\u003eAUC, area under the curve; CI, confidence interval; MPE, max peak enhancement.\u003c/p\u003e\n \u003cp\u003eA nomogram was produced by incorporating the above five independent predictors (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e). It showed good discrimination with an area under the curve (AUC) of 0.857 (95% CI, 0.812\u0026ndash;0.894) (Table \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e and Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eA). The calibration curve showed good agreement between the nomogram-estimated probability of ALNM and the actual ALNM rate in the training cohort, with a mean absolute error of 0.021 (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eD and Table \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003e\u003c/p\u003e\u0026nbsp;\u003ctable id=\"Tab5\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eThe model performance in estimating the risk of ALNM in patients with breast cancer.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eParameters\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eTraining cohort\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eTest cohort\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCut-off value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eN/A\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAUC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.857 (0.812\u0026ndash;0.894)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.858 (0.758\u0026ndash;0.928)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSensitivity (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e78.21 (67.4\u0026ndash;86.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e94.74 (74.0\u0026ndash;99.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSpecificity (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e82.27 (76.6\u0026ndash;87.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e71.43 (57.8\u0026ndash;82.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAccuracy (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e81.21\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e77.33\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePPV (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e61.0 (53.5\u0026ndash;68.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e52.9 (42.3\u0026ndash;63.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNPV (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e91.4 (87.4\u0026ndash;94.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e97.6 (85.5\u0026ndash;99.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePLR (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4.41 (3.2-6.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3.32 (2.2\u0026ndash;5.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNLR (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.26 (0.2\u0026ndash;0.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.074 (0.01\u0026ndash;0.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003ctfoot\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"3\"\u003eNote: Data in parentheses is 95% CI.\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tfoot\u003e\n \u003c/table\u003e\n \u003cp\u003e\u003c/p\u003e\n \u003cp\u003eALNM, axillary lymph node metastasis; AUC, area under the curve; PPV, positive predictive value; NPV, negative predictive value; PLR, positive likelihood ratio; NLR, negative likelihood ratio; CI, confidence interval.\u003c/p\u003e\n \u003cp\u003e\u003c/p\u003e\u0026nbsp;\u003ctable id=\"Tab6\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 6\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eThe calibration efficiency of the prediction nomogram in the training and test cohorts.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eMAE\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eMSE\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e0.9 QoAB\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTraining cohort (n\u0026thinsp;=\u0026thinsp;298)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.021\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e6e-04\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.037\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTest cohort (n\u0026thinsp;=\u0026thinsp;75)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.056\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.00767\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.127\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003cp\u003e\u003c/p\u003e\n \u003cp\u003eMAE, mean absolute error; MSE, mean squared error; QoAB, quantile of absolute error.\u003c/p\u003e\n \u003cp\u003eIn addition, in the training cohort, the cutoff value of 0.3 was selected to distinguish the presence of ALNM, with a sensitivity of 78.21%, specificity of 82.37%, accuracy of 81.21%, positive predictive value (PPV) of 61.0%, negative predictive value (NPV) of 91.4%, positive likelihood ratio (PLR) of 4.41, and negative likelihood ratio (NLR) of 0.26 (Table \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\n \u003ch2\u003eRisk stratification system of the nomogram\u003c/h2\u003e\n \u003cp\u003eIn the training cohort, the cut-off value of 0.3 was selected to distinguish the presence of ALNM (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e). The unsupervised clustering K-Mean algorithm was used to risk stratify the model based on the scores from the nomogram, specifically: 0\u0026ndash;78.874 points for the low-risk group, 92.929\u0026ndash;166.74 points for the medium-risk group, and 178.874\u0026ndash;318.965 points for the high-risk group.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\n \u003ch2\u003ePrediction model validation\u003c/h2\u003e\n \u003cp\u003eGood discrimination with an AUC of 0.858 (95% CI, 0.758\u0026ndash;0.928) was achieved in the test cohort (Table \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e and Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eB). Good calibration was also confirmed, with a mean absolute error of 0.056 (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eD and Table \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eThe decision curve revealed that if the threshold probability of a patient or physician was more significant than 7%, more advantages would be added by using the nomogram to estimate ALNM in breast cancer patients (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eAn example of predicting ALNM (-) correctly and one predicting ALNM(+) correctly were presented, respectively, to illustrate the clinical utility of the constructed nomogram (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e and Fig. \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e). For clinical use, age was determined by drawing a line straight up to the point axis to establish the score associated with the age. Next, this process was repeated for the other four covariates (washout, radiomics score, ratio, and margin). The scores of each covariate were added, and the total score was located on the total score points axis. Last, a line was drawn straight down to the risk of the ALNM axis to obtain the probability. Each parameter had corresponding value (points) that appear in upper toolbar as following: \u0026ldquo;Age 46 years old\u0026rdquo; = 26 points, \u0026ldquo;Washout\u0026thinsp;\u0026gt;\u0026thinsp;0.18\u0026rdquo; = 67 points, \u0026ldquo;Radiomics score\u0026thinsp;\u0026gt;\u0026thinsp;0.2506\u0026rdquo; = 53 points, \u0026ldquo;Ratio\u0026thinsp;\u0026gt;\u0026thinsp;38\u0026rdquo; = 74 points, and \u0026ldquo;Margin was with glitch\u0026rdquo; = 100 points. A summarized total was applied on the bottom scale to obtain the probability of ALNM. Any probability greater than 0.3 (about 176 points) was compatible with ALNM.\u003c/p\u003e\n \u003cp\u003eALNM, axillary lymph node metastasis.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn the current study, we investigated the feasibility and accuracy of the radiomics-based prediction model for prediction of ALNM in patients with breast cancer based on DCE-MRI images of the primary tumor. Our study has three significant findings. First, the radiomics nomogram, based on age, margin, ratio, washout, and radiomics score, showed a favorable ability to discriminate between ALNM (+) and ALNM (-), with AUC values of 0.857 and 0.858 in the training and test cohorts, respectively. Second, a higher degree of kinetic heterogeneity was associated with ALNM. The multivariate analysis showed that higher values of kinetic heterogeneity (ratio [OR 6.26351] and washout [OR 5.30904]), as determined with the nomogram, were associated with ALNM in women with breast cancer. Third, in the stratification of ALNM risk in women with breast cancer and thus allow therapies to be tailored based on individual risk level. The omission of axillary lymph node dissection might be justified in this subgroup of women with low-risk group.\u003c/p\u003e \u003cp\u003eBreast cancer is a heterogeneous tumor with intra-tumoral temporal and spatial variation in cellularity, angiogenesis, and extravascular extracellular matrix [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. It's a truism that the intra-tumor heterogeneity is associated with metastasis and poor prognosis due to inherently invasive biological behavior. However, it remains a challenge to quantify intra-tumoral heterogeneity in a noninvasive way before surgery. Tumor angiogenesis is one of the prerequisites for tumor progression and metastasis, and affects the uptake of contrast media within a tumor during DCE MRI. Therefore, it seems plausible that there are associations between ALNM and the breast cancer kinetic features extracted from DCE-MRI. In the current study, we hypothesized that intra-tumoral heterogeneity might be reflected in tumor enhancement kinetics from DCE-MRI and the concrete characteristic could be quantified with B.K. software. We presumed that higher tumor kinetic heterogeneity as observed on DCE-MRI images might reflect highly heterogeneous tumors with temporal and spatial variation in angiogenesis and various histopathologic components, further leading to the occurrence of ALNM.\u003c/p\u003e \u003cp\u003eVascular endothelial growth factor (VEGF), as we all known, is a major stimulator of angiogenesis, which is frequently overexpressed in breast cancer [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Thus, the various histopathological components in angiogenesis may be reflected in the tumor enhancement kinetics features, which explains the higher kinetic heterogeneity observed in the present study. The new capillaries formed via angiogenesis are typically immature and more permeable than the normal vasculature, which may cause ALNM[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. This could explain the higher kinetic heterogeneity value for the possibility of ALNM.\u003c/p\u003e \u003cp\u003eMost notably, the five DCE-MRI-based kinetic heterogeneity collected in this study showed independent predictive value in predicting ALNM, suggesting that ratio and washout could potentially identify the characteristics of intra-tumor heterogeneity driving aggressive tumor behavior. Patients assigned to the ALNM high-risk group showed increased heterogeneity, corroborating the hypothesis that tumor heterogeneity is related to aggressive tumor behavior. On the other hand, max peak enhancement, persistent, and plateau were not retained in the final model, which suggesting that DCE-MRI-based intra-tumor heterogeneity does not equal pathologic intra-tumor heterogeneity but can be used in prediction models as a valuable parameter.\u003c/p\u003e \u003cp\u003eThe radiomics score was another independently risk factor for predicting ALNM. Establishing a radiomics score with LASSO has demonstrated excellent results in predicting lymph node metastasis in papillary thyroid carcinoma [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e], cervical cancer [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e], pancreatic carcinoma [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e], rectal cancer [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e], and lung cancer [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. Radiomics characteristics are closely related to the microstructure and biological behavior of the tumors [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. The radiomics score is based on the high-dimensional and statistical features, which were extracted from primary breast tumors. In this study, two radiomics features were used to calculate the radiomics score. These features represent the texture information of tumors, which is highly associated with tumor heterogeneity [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eAge and spicule also showed a significant relationship with ALNM, which is consistent with previous research [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e].The rapid tumor growth rate exceeds the nutritional capacity of its blood supply, which is a possible reason of tumor glitch. A glitch is an invasion of the cell edge caused by factors external to the cell, which is often associated with rapid tumor growth and metastasis of breast cancer.\u003c/p\u003e \u003cp\u003eMore noteworthy was the risk stratification of the nomogram, which increased its clinical utility. For patients with low risk, we did not recommend to perform axillary lymph node dissection, but regular ultrasound follow-up, in order to reduce the surgical wound area and reduce the incidence of postoperative complications. For high-risk patients, we suggested that sentinel lymph node biopsy was not necessary and axillary lymph node dissection could be performed directly, thus reducing the intraoperative waiting time of patients and providing sufficient time for doctors to formulate reasonable preoperative surgical methods.\u003c/p\u003e \u003cp\u003eOur study represents a preliminary test of our underlying hypothesis and has several limitations including a retrospective design.\u003c/p\u003e \u003cp\u003eFirst, this study included a small number of patients from a single institution. Multi-institutional prospective studies would help to verify the results. Additional studies with longer follow-up and a larger population are needed to confirm the clinical utility of the nomogram. Second, the extraction of primary tumor features was semi-automatic segmentation, whether it was DCE-MRI or radiomics, which had errors to a certain extent. In the future, more prepared image acquisitions could be performed with fully automatic segmentation methods that have already been applied[\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. Third, tumor hypoxia can be estimated by hypoxic imaging techniques. The correlation between tumor hypoxia and molecular markers is a potential and interesting research on ALNM. Finally, integrating DCE-MRI-based heterogeneity with tumor histopathology and molecular subtypes holds the promise for better ALNM prediction and improved clinical decision making for individual patients.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eIn conclusion, we have proposed a radiomics-based nomogram for predicting ALNM in patients with breast cancer. Moreover, risk stratification can assist in the formulation of individualized treatment plans for patients in clinical practice. The quantitative assessment of DCE-MRI-based heterogeneity may provide biological information related to the aggressiveness of breast cancer that are important determinants of ALNM, and may potentially allow for more individualized treatment strategies. Furthermore, the advantages of the DCE-MRI-based heterogeneity assessment in this study are its noninvasive nature and the potential for easy application in routine clinical practice.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eAUC Area under the curve\u003c/p\u003e\n\u003cp\u003eALN Axillary lymph node\u003c/p\u003e\n\u003cp\u003eALND Axillary lymph node dissection\u003c/p\u003e\n\u003cp\u003eALNM Axillary lymph node metastasis\u003c/p\u003e\n\u003cp\u003eCI Confidence interval\u003c/p\u003e\n\u003cp\u003eDCE-MRI Dynamic contrast-enhanced magnetic resonance imaging\u003c/p\u003e\n\u003cp\u003eDICOM Digital imaging and communications in medicine\u003c/p\u003e\n\u003cp\u003eER Estrogen receptor\u003c/p\u003e\n\u003cp\u003eGLCM Gray level co-occurrence matrix\u003c/p\u003e\n\u003cp\u003eGLSZM Gray level size zone matrix\u003c/p\u003e\n\u003cp\u003eHER 2 Human epidermal growth factor receptor 2\u003c/p\u003e\n\u003cp\u003eLASSO Least absolute shrinkage and selection operator\u003c/p\u003e\n\u003cp\u003eLVI Lymphovascular invasion\u003c/p\u003e\n\u003cp\u003eMAE Mean absolute error\u003c/p\u003e\n\u003cp\u003eMPE Max peak enhancement\u003c/p\u003e\n\u003cp\u003eMRI Magnetic resonance imaging\u003c/p\u003e\n\u003cp\u003eMSE Mean squared error\u003c/p\u003e\n\u003cp\u003eNLR Negative likelihood ratio\u003c/p\u003e\n\u003cp\u003eNPV Negative predictive value\u003c/p\u003e\n\u003cp\u003eOR Odds ratio\u003c/p\u003e\n\u003cp\u003ePLR Positive likelihood ratio\u003c/p\u003e\n\u003cp\u003ePPV Positive predictive value\u003c/p\u003e\n\u003cp\u003ePR Progesterone receptor \u003c/p\u003e\n\u003cp\u003eQoAB Quantile of absolute error\u003c/p\u003e\n\u003cp\u003eROC Receiver operating characteristic\u003c/p\u003e\n\u003cp\u003eROI Regions of interest\u003c/p\u003e\n\u003cp\u003eVEGF Vascular endothelial growth factor\u003c/p\u003e\n\u003cp\u003eVIF Variance inflation factor\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was approved by the institutional review board of Jilin Cancer Hospital (IRB No.2023-001-01) and informed consent was waivered due to retrospective nature.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll authors have reviewed the final version of the paper and would like to take public responsibility for its content.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was supported by Science and Technology Developing Plan of Jilin Province (No.20230203093SF) and Health Technology Capability Enhancement Project of Jilin Province (No.2022JC022).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026apos; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSS conceived and designed this study. TS and DY carried out the collection and assembly of data and drafted the manuscript. MY, JY and HZ were responsible for feature extraction and statistical work. All authors reviewed the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026apos; information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDepartment of Radiology, Jilin Cancer Hospital, Changchun, Jilin, China\u003c/p\u003e\n\u003cp\u003eTongxu Shen, Dingli Ye, Ming Yao, Jieqiong Yan, Han Zhang \u0026amp; Shuangyan Sun\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eSiegel RL, Miller KD, Fuchs HE, Jemal A. Cancer Statistics. Cancer J Clin. 2021;71(1):7\u0026ndash;33.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSamiei S, Granzier RWY, Ibrahim A et al. Dedicated Axillary MRI-Based Radiomics Analysis for the Prediction of Axillary Lymph Node Metastasis in Breast Cancer. Cancers. 2021; 13(4).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang X, Yang Z, Cui W, et al. Preoperative prediction of axillary sentinel lymph node burden with multiparametric MRI-based radiomics nomogram in early-stage breast cancer. Eur Radiol. 2021;31(8):5924\u0026ndash;39.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHan L, Zhu Y, Liu Z, et al. Radiomic nomogram for prediction of axillary lymph node metastasis in breast cancer. Eur Radiol. 2019;29(7):3820\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim JY, Kim JJ, Hwangbo L, et al. Kinetic Heterogeneity of Breast Cancer Determined Using Computer-aided Diagnosis of Preoperative MRI Scans: Relationship to Distant Metastasis-Free Survival. Radiology. 2020;295(3):517\u0026ndash;26.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFan M, Zhang P, Wang Y, et al. Radiomic analysis of imaging heterogeneity in tumours and the surrounding parenchyma based on unsupervised decomposition of DCE-MRI for predicting molecular subtypes of breast cancer. Eur Radiol. 2019;29(8):4456\u0026ndash;67.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim JJ, Kim JY, Suh HB, et al. Characterization of breast cancer subtypes based on quantitative assessment of intratumoral heterogeneity using dynamic contrast-enhanced and diffusion-weighted magnetic resonance imaging. Eur Radiol. 2021. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00330-021-08166-4\u003c/span\u003e\u003cspan address=\"10.1007/s00330-021-08166-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao R, Ma WJ, Tang J, et al. Heterogeneity of enhancement kinetics in dynamic contrast-enhanced MRI and implication of distant metastasis in invasive breast cancer. Clin Radiol. 2020;75(12):961. e925-961 e932.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChitalia RD, Rowland J, McDonald ES, et al. Imaging Phenotypes of Breast Cancer Heterogeneity in Preoperative Breast Dynamic Contrast Enhanced Magnetic Resonance Imaging (DCE-MRI) Scans Predict 10-Year Recurrence. Clin cancer research: official J Am Association Cancer Res. 2020;26(4):862\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCollins GS, Reitsma JB, Altman DG, Moons KGM, members of the Tg. Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD): The TRIPOD Statement. Eur Urol. 2015;67(6):1142\u0026ndash;51.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBitencourt AGV, Gibbs P, Rossi Saccarelli C, et al. MRI-based machine learning radiomics can predict HER2 expression level and pathologic response after neoadjuvant therapy in HER2 overexpressing breast cancer. EBioMedicine. 2020;61:103042.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee JY, Lee KS, Seo BK et al. Radiomic machine learning for predicting prognostic biomarkers and molecular subtypes of breast cancer using tumor heterogeneity and angiogenesis properties on MRI. Eur Radiol. 2021; 1\u0026ndash;11.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMalekan M, Ebrahimzadeh MA. Vascular Endothelial Growth Factor Receptors [VEGFR] as Target in Breast Cancer Treatment: Current Status in Preclinical and Clinical Studies and Future Directions. Curr Top Med Chem. 2022;22(11):891\u0026ndash;920.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSiemann DW, Chaplin DJ, Horsman MR. Realizing the Potential of Vascular Targeted Therapy: The Rationale for Combining Vascular Disrupting Agents and Anti-Angiogenic Agents to Treat Cancer. Cancer Invest. 2017; 14;35(8):519\u0026ndash;534.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang M, Li C, Tang S, et al. Nomogram Based on Shear-Wave Elastography Radiomics Can Improve Preoperative Cervical Lymph Node Staging for Papillary Thyroid Carcinoma. Thyroid: official J Am Thyroid Association. 2020;30(6):885\u0026ndash;97.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJin X, Ai Y, Zhang J, et al. Noninvasive prediction of lymph node status for patients with early-stage cervical cancer based on radiomics features from ultrasound images. Eur Radiol. 2020;30(7):4117\u0026ndash;24.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBian Y, Guo S, Jiang H, et al. Relationship Between Radiomics and Risk of Lymph Node Metastasis in Pancreatic Ductal Adenocarcinoma. Pancreas. 2019;48(9):1195\u0026ndash;203.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNakanishi R, Akiyoshi T, Toda S, et al. Radiomics Approach Outperforms Diameter Criteria for Predicting Pathological Lateral Lymph Node Metastasis After Neoadjuvant (Chemo)Radiotherapy in Advanced Low Rectal Cancer. Ann Surg Oncol. 2020;27(11):4273\u0026ndash;83.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXie Y, Zhao H, Guo Y, et al. A PET/CT nomogram incorporating SUVmax and CT radiomics for preoperative nodal staging in non-small cell lung cancer. Eur Radiol. 2021;31(8):6030\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHu HT, Wang Z, Huang XW, et al. Ultrasound-based radiomics score: a potential biomarker for the prediction of microvascular invasion in hepatocellular carcinoma. Eur Radiol. 2019;29(6):2890\u0026ndash;901.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiao H, Jiang X, Pang Z et al. Deep Convolutional Neural Networks-Based Automatic Breast Segmentation and Mass Detection in DCE-MRI. Comput Math Methods Med. 2020; 2020:2413706.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRahimpour M, Saint Martin MJ, Frouin F, et al. Visual ensemble selection of deep convolutional neural networks for 3D segmentation of breast tumors on dynamic contrast enhanced MRI. Eur Radiol. 2023;33(2):959\u0026ndash;69.\u003c/span\u003e\u003c/li\u003e\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":"Kinetic heterogeneity, axillary lymph node metastasis, breast cancer, dynamic contrast-enhanced magnetic resonance imaging, radiomics nomogram","lastPublishedDoi":"10.21203/rs.3.rs-4445164/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4445164/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eTo investigate whether kinetic heterogeneity, assessed via dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI)-based radiomics nomogram, associated with axillary lymph node metastasis (ALNM) in cN0 breast cancer.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003e373 consecutive women with cN0 breast cancer who underwent preoperative DCE-MRI were retrospectively evaluated from 2016 to 2020. The kinetic heterogeneity (a measure of heterogeneity in the proportions of peak enhancement, peak enhancement ratio, persistent, plateau, and washout) was assessed with DCE-MRI using B.K. software automatically. Radiomics features were extracted from magnetic resonance imaging (MRI) images of the primary breast cancer lesion. The minimum redundancy maximum relevance algorithm was used to select ALNM positively-related features and radiomics score was constructed. Clinical features, MRI features, kinetic heterogeneity, and radiomics score were screened out by multivariate logistic regression analysis, and the nomogram was constructed from these characteristics. Possible associations between DCE-MRI-based kinetic heterogeneity and ALNM were analyzed. The unsupervised clustering K-Mean algorithm was use to risk stratification.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eFive independent risk factors were screened out to build the nomogram, including: age, margin, ratio, washout, and radiomics score. The area under the receiver operating characteristic curve was 0.857 and 0.858 in the training and test cohorts, respectively. The risk stratification system divided all patients into three risk groups. Axillary lymph node dissection was not recommended for the low-risk group and was strongly recommended for the high-risk group.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eRadiomic analysis of kinetic heterogeneity based on the DCE-MRI images has the potential to more accurately identify tumor kinetic features and serve as a valuable clinical marker to enhance the prediction of ALNM in cN0 breast cancer.\u003c/p\u003e","manuscriptTitle":"Kinetic heterogeneity is associated with axillary lymph node metastasis in cN0 breast cancer based on dynamic contrast-enhanced magnetic resonance imaging radiomics nomogram","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-06-04 19:01:10","doi":"10.21203/rs.3.rs-4445164/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":"0a8d096d-6e93-46e1-bfd5-f245812154c5","owner":[],"postedDate":"June 4th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-05-29T01:58:05+00:00","versionOfRecord":[],"versionCreatedAt":"2024-06-04 19:01:10","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4445164","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4445164","identity":"rs-4445164","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}