Postoperative One Year Prediction for Patients with Cervical Spinal Cord Injury Based on Deep Learning and Radiomics

Postoperative One Year Prediction for Patients with Cervical Spinal Cord Injury Based on Deep Learning and Radiomics | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Postoperative One Year Prediction for Patients with Cervical Spinal Cord Injury Based on Deep Learning and Radiomics Fabin Lin, Kaifeng Wang, Ruxian Wang, Yang Wu, Chunmei Chen, Yongjiang Wang, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4848654/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: Cervical spinal cord injury (SCI) can lead to significant impairments, requiring extensive care and posing considerable challenges in predicting postoperative outcomes. This study aimed to develop and validate a deep learning radiomics (DLR) model combining deep learning and radiomics features to improve the prognostic prediction of cervical SCI. Methods: This retrospective study included 82 patients with confirmed cervical SCI from three hospitals, collected between January 2012 and January 2021. Patients were divided into good prognosis and poor prognosis groups based on postoperative ASIA grade improvement. Preoperative MRI images were processed using various filtering techniques, and regions of interest (ROI) were segmented and analyzed to extract radiomics features. Deep learning models (ResNet-18, ResNet-50, and ResNet-101) were trained. Features from both radiomics and deep learning models were combined and selected 、 to build the final predictive model using MLP. Results: ResNet-50 outperformed other models, demonstrating an AUC of 0.8750 in the test set. The combined model (Rad + ResNet-50) showed the highest prognostic value with an AUC of 0.9220 in the test set. Grad-CAM images enhanced the interpretability of the model by highlighting critical areas for prognosis prediction. Conclusion: Integrating deep learning and radiomics features significantly improves the prediction accuracy for cervical SCI outcomes. The Rad + ResNet-50 model, with its superior performance and interpretability, holds promise for clinical applications, offering a robust tool for predicting functional prognosis in cervical SCI patients. Further prospective studies with larger datasets are needed to validate these findings. Health sciences/Neurology/Neurological disorders/Neuromuscular disease Biological sciences/Neuroscience/Diseases of the nervous system/Neurodegeneration Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Cervical spinal cord injury (SCI) can result in a spectrum of outcomes, from complete to incomplete tetraplegia, leading to significant physical, psychological, and social consequences for those affected [ 1 ]. Individuals with cervical SCI experience substantial impairments in various aspects of life, requiring extensive care and effort. Preoperative prediction is crucial for postoperative patients, yet the associated risk factors remain unidentified [ 2 , 3 ]. The convergence of machine learning (ML) and medical advancements has significantly improved diagnosis, risk assessment, and treatment response prediction [ 4 – 7 ]. ML algorithms can identify subtle data patterns often missed by human experts, though research on cervical SCI remains limited [ 8 – 11 ]. Radiomics, which extracts numerous quantitative features from medical images and analyzes them using advanced algorithms, has been employed to detect spinal cord compression at different locations and guide management decisions [ 12 ]. Deep learning (DL), utilizing convolutional neural networks to extract image features, has driven significant advancements in artificial intelligence and human-computer interactions. This approach reduces diagnostic errors and maximizes the potential of radiomics data. Furthermore, deep transfer learning enables the application of deep learning to small datasets. As neuroscience advances intersect with technological innovations, neurosurgery finds unique opportunities to optimize patient care through ANNs. However, integrating clinical relevance with complex, non-linear data for accurate prognostic modeling remains challenging[ 12 ]. This study aimed to compare traditional radiomics with deep learning-enhanced radiomics and to develop and validate a deep learning radiomics (DLR) model for accurately predicting cervical SCI outcomes. Materials and methods Patient Selection and Study Design This retrospective study received approval from the ethical committees of three hospitals (approval no. 2021KY138), with informed consent waived. Data from 223 patients collected between January 2012 and January 2021 were included. Inclusion criteria : Confirmed diagnosis of cervical spinal cord injury (acute injury to nerve structures in the spinal canal causing temporary or permanent sensory and motor function changes, with or without vesicorectal dysfunction) Complete case data and preoperative MRI images No prior treatment before surgery Exclusion criteria Incomplete clinical records or inadequate imaging data. Initially, 223 patients were included: 118 from Ordos Central Hospital, and 19 from The First People’s Hospital of ChangDe City, 86 from Fujian Medical University Union Hospital. We excluded 86 patients with non-cervical spinal cord injuries. Of the remaining 137 patients, 55 were excluded due to incomplete or low-quality MRI images. Ultimately, 82 patients were analyzed, divided into good prognosis (41 patients) and poor prognosis (41 patients). The patient recruitment map is shown in Fig. 1A, and the study design and pipeline are illustrated in Fig. 1C. Surgical Procedures All participants underwent neck surgery via either an anterior or posterior approach, including anterior cervical discectomy, microdecompression of the spinal canal, implant fusion, internal fixation with steel plate, or posterior median cervical internal fixation, implant fusion, and decompression. Based on the study by Yanlin Yin et al., there were no significant differences in postoperative cervical and neurological recovery between anterior and posterior surgical interventions[ 13 ]. Follow-ups were conducted to monitor progress. Detailed descriptions of the surgical techniques are provided below: Patient Position : Prone for the posterior approach, supine for the anterior approach. Anesthesia Method : Tracheal intubation with sedation combined with general anesthesia. Preoperative Localization : Head frame fixation after retraction (posterior approach) and immobilization. Using X-ray front and side fluoroscopy of the C-arm machine, the surgical segments were localized and marked. Localization of Puncture : After routine disinfection and towel spreading, the surgical incision was made according to intraoperative C-arm X-ray positive and lateral fluoroscopy film, with layer-by-layer incision of the skin, subcutaneous tissue, and deep fascia. Surgical Steps [ 13 , 14 ]: Anterior Approach : Vertebral body resection and thorough excision of implicated intervertebral discs. Cervical fusion was achieved using either allograft fibular struts or autografts, secured with semi-rigid locking plates. Postoperatively, cervical immobilization with a collar was maintained for two weeks until wound healing. Posterior Approach : Using a Mayfield clamp or horseshoe headrest, the cervical spine was stabilized in a neutral position. The extent of cervical laminoplasty was predetermined through preoperative imaging. Transverse mass screws were deployed and affixed with connecting rods to achieve segmental decompression by excising the ligamenta flava and laminae of the targeted segments. Postoperative Care : Complete hemostasis was ensured, paravertebral muscles were repositioned, and the subcutaneous layer and skin were sutured. The operative site was dressed with sterile gauze, and a drainage tube was inserted. Patients were encouraged to start range-of-motion exercises immediately after the drainage tube was removed and were discharged once they were able to walk. Both cohorts were managed in alignment with the hospital’s clinical guidelines for cervical spine interventions, incorporating a uniform regimen of physical therapy. Clinical Data Collection and ASIA Scale Assessment Patient demographic and clinical information, including age, gender, and cause of spinal cord injury, were collected. The preoperative and postoperative ASIA scales were utilized to calculate the difference. During postoperative follow-up, the ASIA score criteria remained consistent with preoperative standards. Patients were classified into good prognosis and poor prognosis groups based on whether their ASIA grade improved by at least one level post-surgery: a decrease indicated a good prognosis, otherwise poor prognosis. Regarding the ASIA grade scoring criteria, please refer to Supplementary File 1. Acquisition and Preprocessing of T2-weighted MRI Images The preoperative MRI scans for all enrolled patients, featuring T2-weighted imaging (T2WI), were conducted using 3.0 Tesla MRI technology. The machine settings at different centers are shown in Table 3 . After obtaining the T2-weighted MRI images, preprocessing included normalization to the range of 0–1, resampling to a voxel size of 1x1x1, and processing image noise using multiple filters: Median, Average Gaussian and Box filtering. Each filter outputted as a separate channel rather than being overlaid. These operations aimed at reducing multicenter effects and enhancing the model's generalization ability. Detailed descriptions of each filter can be found in the supplementary file2. ROI Segmentation and Radiomics Analysis Segmentation of Regions of Interest (ROI) The 3D Slicer Software Application v5.0.2 ( https://www.slicer.org/ ) was used to segment the Region of Interest (ROI). The ROI comprised two segments: the spinal cord injury region (segmentation 1) and the unaffected spinal cord regions (segmentation 2). Segmentation 1 is defined as the area of evident spinal cord injury, such as edema and significant spinal cord compression [ 15 – 17 ]. Segmentation 2 is defined as the region excluding the anatomical landmarks of Segmentation 1, including the upper end level aligned with the foramen magnum and the lower end level aligned with the lower edge of the first lumbar vertebra in adults. This defined the extent of the ROI, encompassing the entire spinal cord while carefully excluding vertebral and cerebrospinal fluid regions. This approach ensures that significant biomarkers outside the area of spinal cord injury are not overlooked. Figure 1B illustrates our ROI delineation. Consistency between original images and the four filtering techniques ensured uniformity in ROI delineation for each patient. Each participant's ROIs were independently segmented by two raters, R.W. and Y.W. Intra-rater reliability was assessed by the same radiologist segmenting 30 cases two weeks apart, while inter-rater reliability was evaluated by another radiologist segmenting the same 30 cases. Intraclass correlation coefficients (ICCs) were calculated to assess both intra- and inter-rater reliability, with an ICC > 0.75 indicating strong consistency. Extraction of Radiomics Features Using the Pyradiomics module within 3D Slicer, radiomics features were extracted and organized into seven categories for each segmentation in each filtering technique: (1) Shape-based features: 14, (2) First-order features: 18, (3) Gray-level dependence matrix (GLDM) features: 14, (4) Gray-level size zone matrix (GLSZM) features: 16, (5) Neighboring gray-tone difference matrix (NGTDM) features: 5, (6) Gray-level run-length matrix (GLRLM) features: 16, and (7) Gray-level co-occurrence matrix (GLCM) features: 24. Processing with the four filtering techniques yielded a total of 1,070 radiomics features per patient. Detailed feature categories are provided in Supplementary File 3 and4. Deep learning analysis Training Deep Learning Models For training the deep learning models, the sagittal plane image with the maximum ROI for each patient was selected and resampled to 224x224 pixels. Other image data preparation steps followed the same preprocessing process as described above. Three deep learning models were used: ResNet18, ResNet50, and ResNet101, all pretrained on the ImageNet dataset [ 18 ]. Five different channels were input into each model: the original image, box, average, Gaussian and median filtering. Each channel underwent an independent yet identical training process with consistent parameter settings. The five separately trained channels ultimately converged to contribute to the classification task. Data were split into training and test sets in a 7:3 ratio. We have carefully set the parameters (initial learning rate: 0.01, batch: 64, 50 epochs). The classification targets were good prognosis (1) and poor prognosis (0). Model parameters were divided into two parts: 1. the backbone layer (backbone) and 2. the task-specific layer (task-spec). Task-specific parameters were randomly initialized, while backbone parameters utilized the pre-trained model parameters from ImageNet. For the task-specific parameters, the cosine annealing learning rate decay algorithm was applied [ 19 ], with details provided in the supplementary file5. Deep learning feature extraction The penultimate layer of the model (AveragePooling layer) was used for feature extraction. Our dataset, which includes five categories of images, was independently trained through five identical deep learning channels. Deep learning features were extracted separately for each channel. For ResNet50 and ResNet101 the feature dimensionality was 2048, while for and ResNet18, it was 512. To reduce the risk of overfitting and improve the model's generalization capability, principal component analysis (PCA) was applied to compress the extracted features to 32 dimensions per channel, resulting in 160 deep learning features (32x5) per patient. In multi-center studies, the ComBat compensation procedure addresses differences arising from imaging protocols and various MRI scanners[ 18 , 19 ]. This method shows considerable promise in improving the reproducibility of research conducted across multiple centers. Feature engineering First, We utilized the ComBat compensation procedure on all extracted features to mitigate multi-center effects[ 20 ]. After that, Radiomics features (1,070) were combined with deep learning features (160), resulting in 1,230 fused features per patient. Before feature selection, data were standardized using Z-scores to convert all features to a uniform mean of 0 and a standard deviation of 1. Initial Screening : Features with a p-value < 0.05 were retained using the Mann-Whitney U test. Redundancy Reduction : Features were screened for high repeatability using the Spearman rank correlation coefficient, retaining only one feature from any pair with a correlation coefficient greater than 0.9[ 21 ]. A greedy recursive deletion strategy was applied to remove the feature with the highest redundancy in each iteration. Feature Selection : The Minimum Redundancy Maximum Relevance (MRMR) algorithm was used to select the best features. A 10-fold cross-validated Least Absolute Shrinkage and Selection Operator (LASSO) regression identified features with non-zero coefficients for subsequent modeling. Development and assessment of models The study employed MLP machine learning methods, implemented using the scikit-learn package (version: 0.18) in Python 3.9. The dataset was randomly split into training (70%) and test (30%) datasets to evaluate model accuracy. Five-fold cross-validation and Grid Search were used to determine the optimal hyperparameters for each model. Evaluation metrics included AUC, Decision Curve Analysis (DCA), Calibration curve, specificity, sensitivity, and accuracy, among others. Statistical analysis Categorical variables and normally distributed variables were analyzed using the chi-square test and t-tests, respectively. A p-value < 0.05 indicated statistical significance. Python (version 3.9, http://www.python.org ) was utilized for all analyses. Results Patient characteristics A total of 82 patients Cervical SCI were enrolled in this study, who were aged from 9 to 84 years (a mean age of 52 years). As shown in the methods section. The entire dataset was divided into two distinct groups: good prognosis and bad prognosis. Table 1 summarizes the comparison of demographic characteristics of the two groups. Table 1 Clinical characteristics Feature_name Total Difference in ASI number, no improvement(n = 41) Difference in ASI number, improvement(n = 41) pvalue Age 52.33 ± 12.32 52.83 ± 11.01 51.83 ± 13.63 0.593625 Gender 0.567806 Female 15(18.29) 9(21.95) 6(14.63) Male 67(81.71) 32(78.05) 35(85.37) Injury mechanism: 0.082036 Fall on the same level 18(21.95) 13(31.71) 5(12.20) Fall from a height less than 1 meter 2(2.44) 1(2.44) 1(2.44) Fall from a height greater than 1 meter 37(45.12) 14(34.15) 23(56.10) Traffic accident 17(20.73) 7(17.07) 10(24.39) Others (e.g., sports) 8(9.76) 6(14.63) 2(4.88) Number of injured segments 0.452569 1 22(26.83) 13(31.71) 9(21.95) 2 36(43.90) 16(39.02) 20(48.78) 3 14(17.07) 6(14.63) 8(19.51) 4 7(8.54) 5(12.20) 2(4.88) 5 1(1.22) 1(2.44) 0 6 1(1.22) 0 1(2.44) 7 1(1.22) 0 1(2.44) Fracture 0.03609 No 54(65.85) 32(78.05) 22(53.66) Yes 28(34.15) 9(21.95) 19(46.34) Preoperative_ASIA_grade < 0.001 Grade A 13(15.48) 6(14.63) 7(16.28) Grade B 15(17.86) 4(9.76) 11(25.58) Grade C 24(28.57) 4(9.76) 20(46.51) Grade D 26(30.95) 21(51.22) 5(11.63) Grade E 6(7.14) 6(14.63) 0 Postoperative_ASIA_grade 0.0275 Grade A 7(8.33) 7(17.07) 0 Grade B 7(8.33) 3(7.32) 4(9.30) Grade C 12(14.29) 8(19.51) 4(9.30) Grade D 48(57.14) 19(46.34) 29(67.44) Grade E 10(11.90) 4(9.76) 6(13.95) Forecasting with the Deep earning and radiomics model Radiomics features and modeling A total of 1045 features passed the ICC test. After extensive feature screening, seven features were incorporated into the radiomics model. The model demonstrated an AUC of 0.869 (95% CI 0.7862–0.9525) in the training cohort, with sensitivity (0.875) and specificity (0.727), respectively. In the test cohort, the AUC was 0.7780 (95% CI 0.5285-1), with sensitivity (0.778) and specificity (0.750). Detailed statistical results and curves are presented in Fig. 2 and Table 2 . Table 2 Comparision results of performance different Deep learning models Model Cohort AUC 95% CI Accuracy Sen Spe PPV NPV Pre Recall F1 Rad train 0.8690 0.7862–0.9525 0.8000 0.8750 0.7270 0.7570 0.8570 0.7570 0.8750 0.8120 test 0.7780 0.5285–1.0000 0.7650 0.7780 0.7500 0.7780 0.7500 0.7780 0.7780 0.7780 ResNet18 train 0.9480 0.9014–0.9945 0.8770 0.8480 0.9060 0.9030 0.8530 0.9030 0.8480 0.8750 test 0.7920 0.5695–1.0000 0.7650 0.8750 0.6670 0.7000 0.8570 0.7000 0.8750 0.7780 Rad+ ResNet18 train 0.8570 0.7703–0.9437 0.7690 0.6250 0.9090 0.8700 0.7140 0.8700 0.6250 0.7270 test 0.8470 0.6474–1.0000 0.8240 0.7780 0.8750 0.8750 0.7780 0.8750 0.7780 0.8240 ResNet50 train 0.9750 0.9358–1.0000 0.9550 0.9390 0.9700 0.9690 0.9410 0.9690 0.9390 0.9540 test 0.8750 0.6898–1.0000 0.8750 0.8750 0.8750 0.8750 0.8750 0.8750 0.8750 0.8750 Rad+ ResNet50 train 0.9400 0.8835–0.9971 0.9240 0.9090 0.9390 0.9370 0.9120 0.9370 0.9090 0.9230 test 0.9220 0.7919–1.0000 0.8750 0.8750 0.8750 0.8750 0.8750 0.8750 0.8750 0.8750 ResNet101 train 0.9270 0.8682–0.9849 0.8640 0.8790 0.8480 0.8530 0.8750 0.8530 0.8790 0.8660 test 0.8590 0.6692–1.0000 0.8120 0.8750 0.7500 0.7780 0.8570 0.7780 0.8750 0.8240 Rad+ ResNet101 train 0.8820 0.8030–0.9602 0.8150 0.8790 0.7500 0.7840 0.8570 0.7840 0.8790 0.8290 Table 3 MRI Scanners in Three Centers Item Fujian Medical University Union Hospital Ordos Central Hospital The First People’s Hospital of ChangDe City MRI Machine GE Signa GE Signa GE Signa TR Settings (ms) 2500–3000 2100 2000 TE Settings (ms) 100–110 120 80 Slice Thickness (mm) 3 3 3 Slice Spacing (mm) 1 1 1 Matrix Configuration 256 × 512 256 × 512 256 × 512 Field of View (FOV) (mm × mm) 240 × 240 240 × 240 240 × 240 Imaging Plane Sagittal Sagittal Sagittal Deep learning model The features used to build ResNet-18, ResNet-50, and ResNet-101 were 10, 9, and 10, respectively. With regard to the predictions generated by the deep learning models, among the three models we utilized, ResNet50 performed the best, followed by ResNet101. In the test set, the AUCs for ResNet50, ResNet101, and ResNet18 were 0.8750, 0.8590, and 0.7920, respectively; their accuracies were 0.8750, 0.8120, and 0.7650; sensitivities were 0.8750, 0.8750, and 0.8750; specificities were 0.8750, 0.7500, and 0.6670. (Fig. 2, Table 2 ). Supplementary File 6 illustrates the Lasso regression paths and Mean Squared Error (MSE) plots for each model. Supplementary File 7 provides the feature weights utilized in constructing these models. Combined model Considering the number of features used to establish each model, the radiomics + ResNet18 and radiomics + ResNet50 models embraced the inclusion of nine features(Figure4), while the radiomics + ResNet101 model was constructed using eight select features. Among these, the Rad + ResNet50 model also performed the best: In the test set, the AUCs for Rad + ResNet50, Rad + ResNet101, and Rad + ResNet18 were 0.9220, 0.8750, and 0.8470, respectively; their accuracies were 0.8750, 0.8820, and 0.8240; sensitivities were 0.8750, 0.8750, and 0.7780; specificities were 0.8750, 0.8890, and 0.8750(Fig. 2). For more details, please reference Table 2 . DCA provided additional supporting evidence for these conclusions. DCA shows that when the threshold probability surpasses the 20% mark, the model produces more net benefit in the current study. The Decision Curve Analysis (DCA) and calibration plots indicate strong clinical utility. (Fig. 3A-B) Grad-CAM Visualizing the final convolutional layer using Grad-CAM is an effective method to enhance model interpretability. As illustrated in Fig. 3C, the darker blue regions highlight crucial areas for prognosis prediction. These regions are primarily focused on the edematous spinal cord, deformed vertebrae, and swollen ligaments. Discussion In this study, we included 82 patients with cervical spinal cord injury (SCI). We discovered that the predictive capability of the feature fusion model was enhanced compared to using radiomics or CNNs individually. Ultimately, we found that the Rad + ResNet-50 model had the highest prognostic predictive value for cervical SCI, with AUCs of 0.940 and 0.922 in the training and test cohorts, respectively. Contributions of This Study and Comparison with Previous Studies Functional prognosis after cervical spinal cord injury (SCI), particularly regarding regaining functional independence, is a primary focus in rehabilitation. Predicting SCI recovery can be challenging, but the ASIA grade is a key benchmark for assessing clinical recovery and long-term prognosis [ 22 ]. In our study, 82 cervical SCI patients were divided into two groups based on whether their ASIA index improved after one year. Table 1 shows preoperative and postoperative ASIA grades with p < 0.05, indicating clinical significance. Yann Facchinello et al. also highlight that the severity of the neurological deficit at admission, indicated by the ASIA grade, is a critical predictor of recovery [ 23 ]. In addition to the ASIA score, MRI images are crucial for prognosis, with metrics shown to be reproducible and predictive of clinical outcomes[ 24 ].Hyun-Joon Yoo et al. used machine learning algorithms to create a prediction model for SCI patients based solely on clinical variables, achieving promising results [ 25 ]. However, their approach did not include the prognostic value of imaging, particularly MRI. The BASIC (Brain and Spinal Injury Center) score, developed to predict neurological improvement based on high signal intensity on axial T2-weighted MRI, has proven effective [ 25 ] [ 10 ]. However, this score focused on a limited set of manually measured MRI features, such as injury length and spinal canal compromise, overlooking many other potential features. Our study leverages deep learning radiomics to extract and model MRI features. This approach reduces the risk of missing significant features, as evidenced by the combined model (Rad + ResNet50 AUC: 0.9220), which demonstrates improved performance. We noted that in another study using deep learning, Okimatsu and colleagues developed a CNN model to quantify radiographic characteristics and predict 1-month neurological outcomes in acute SCI patients, achieving 71% accuracy[ 26 ]. We extended the follow-up period (1-year) and also improved the model's accuracy to 92.2%. Integrating handcrafted radiomics (HCR) and deep transfer learning (DTL) features for cervical SCI prognosis is novel, and our approach uses standard image data without special training, highlighting its potential. Training deep learning models on small datasets has limitations, though deep transfer learning can help. André Wirries suggested that using small datasets to train deep learning models for clinical applications is practical[ 27 ]. We enhanced model performance through random shifts, rotations, and horizontal flips during training to increase data diversity[ 28 ]. Expanding the dataset and controlling feature numbers further improved results. We implemented rigorous feature selection processes in our experiments[ 29 , 30 ]. ResNet’s structure uses shortcut connections, enabling effective fusion training[ 31 ]. To evaluate different models, we tested ResNet-18, ResNet-50, and ResNet-101. Both deep learning and combined models showed that ResNet-50 performed best. Despite having fewer layers and higher Top-1 and Top-5 error rates than ResNet-101[ 31 ], ResNet-50’s moderate depth proved advantageous in this research. This indicates that while deeper networks may offer superior learning capabilities, selecting an appropriate network depth for specific research can lead to better model fitting and other benefits, such as reduced computational resources and time. Additionally, ResNet-18 underperformed compared to ResNet-50, suggesting lower-dimensional ResNet networks may lack sufficient fitting capacity. Deep Learning Model Interpretability Although deep learning shows promise in disease diagnosis and prognosis, the interpretability of these models can hinder their broader application[ 32 , 33 ]. Some studies use post-hoc methods or supervised machine learning models to interpret deep learning algorithm outputs[ 34 ]. For example, Yixin Wang visualized SVM features using a SHAP-based method[ 34 ]. In our study, we generated Grad-CAM images to enhance model interpretability, similar to how Kim, Y. and colleagues used heatmaps to identify regions crucial for prognosis prediction[ 35 ]. limitations There are some limitations to this study that further studied and explored in future work. First, the sample size of cervical SCI was small. Although our findings reflect the predictive ability of deep learning features to a certain extent, more data would be more convincing. Second, this was a restrospective study, more prospective data are needed to verify the effectiveness of the model. Conclusions In conclusion, the DLR features were fused with cervical SCI features based on MRI images, which improved the identification ability of a single radiomics prognostic model for cervical SCI. And we developed a new prognostic model (Rad + ResNet-50) which had the highest predictive value for prognosis of cervical SCI with AUCs of 0.940 and 0.922 in the training cohort and test cohort, respectively. Declarations Funding/Support: This research received no external funding Conflict of Interest: The authors have nothing to disclose. Research Ethics: Institutional Review Board approval was obtained. The study was conducted according to the guidelines of the Declaration of Helsinki, and was approved by Institutional Review Board (approval no. 2021KY138). Acknowledgments: We thank the colleagues in our department for their help in our study. Informed Consent Statement: Written informed consent was not required for this study because this is a retrospective study. Data availability ：The datasets used and/or analysed during the current study available from the corresponding author on reasonable request. References Alizadeh, A., S.M. Dyck, and S. Karimi-Abdolrezaee, Traumatic Spinal Cord Injury: An Overview of Pathophysiology, Models and Acute Injury Mechanisms . Front Neurol, 2019. 10: p. 282. Aarabi, B., et al., Intramedullary Lesion Length on Postoperative Magnetic Resonance Imaging is a Strong Predictor of ASIA Impairment Scale Grade Conversion Following Decompressive Surgery in Cervical Spinal Cord Injury . Neurosurgery, 2017. 80(4): p. 610–620. Wichmann, T.O., et al., Early clinical predictors of functional recovery following traumatic spinal cord injury: a population-based study of 143 patients . Acta Neurochirurgica, 2021. 163(8): p. 2289–2296. Giger, M.L., Machine Learning in Medical Imaging. J Am Coll Radiol, 2018. 15(3 Pt B): p. 512–520. Wong, K.K.L., L. Wang, and D. Wang, Recent developments in machine learning for medical imaging applications . Comput Med Imaging Graph, 2017. 57: p. 1–3. English, M., et al., Machine Learning in Neuro-Oncology, Epilepsy, Alzheimer's Disease, and Schizophrenia . Acta Neurochir Suppl, 2022. 134: p. 349–361. Dietz, N., et al., Evaluation of Predictive Models for Complications following Spinal Surgery . J Neurol Surg A Cent Eur Neurosurg, 2020. 81(6): p. 535–545. Inoue, T., et al., XGBoost, a Machine Learning Method, Predicts Neurological Recovery in Patients with Cervical Spinal Cord Injury . Neurotrauma Rep, 2020. 1(1): p. 8–16. Dietz, N., et al., Machine learning in clinical diagnosis, prognostication, and management of acute traumatic spinal cord injury (SCI): A systematic review . J Clin Orthop Trauma, 2022. 35: p. 102046. Haefeli, J., et al., Multivariate Analysis of MRI Biomarkers for Predicting Neurologic Impairment in Cervical Spinal Cord Injury . AJNR Am J Neuroradiol, 2017. 38(3): p. 648–655. Khan, O., et al., Use of Machine Learning and Artificial Intelligence to Drive Personalized Medicine Approaches for Spine Care . World Neurosurg, 2020. 140: p. 512–518. Azimi, P., et al., A Review on the Use of Artificial Intelligence in Spinal Diseases . Asian Spine J, 2020. 14(4): p. 543–571. Yin, Y., et al., The Efficacy of Anterior Cervical Corpectomy and Fusion and Posterior Total Laminectomy on Cervical Spinal Cord Injury and Quality of Life. Comput Math Methods Med, 2022. 2022: p. 8216339. Seng, C., et al., Surgically treated cervical myelopathy: a functional outcome comparison study between multilevel anterior cervical decompression fusion with instrumentation and posterior laminoplasty . Spine J, 2013. 13(7): p. 723–31. Tarawneh, A.M., et al., Can MRI findings predict the outcome of cervical spinal cord Injury? a systematic review . European Spine Journal, 2020. 29(10): p. 2457–2464. Zhang, M.Z., et al., Predicting postoperative recovery in cervical spondylotic myelopathy: construction and interpretation of T2(*)-weighted radiomic-based extra trees models . Eur Radiol, 2022. 32(5): p. 3565–3575. Burns, J.E., J. Yao, and R.M. Summers, Vertebral Body Compression Fractures and Bone Density: Automated Detection and Classification on CT Images . Radiology, 2017. 284(3): p. 788–797. Orlhac, F., et al., Validation of A Method to Compensate Multicenter Effects Affecting CT Radiomics . Radiology, 2019. 291(1): p. 53–59. Lucia, F., et al., External validation of a combined PET and MRI radiomics model for prediction of recurrence in cervical cancer patients treated with chemoradiotherapy . Eur J Nucl Med Mol Imaging, 2019. 46(4): p. 864–877. Bi, S., et al., Multi-parametric MRI-based radiomics signature for preoperative prediction of Ki-67 proliferation status in sinonasal malignancies: a two-centre study . Eur Radiol, 2022. Wang, W., et al., Development and Validation of a Computed Tomography-Based Radiomics Signature to Predict Response to Neoadjuvant Chemotherapy for Locally Advanced Gastric Cancer . JAMA Netw Open, 2021. 4(8): p. e2121143. van Middendorp, J.J., et al., Diagnosis and prognosis of traumatic spinal cord injury . Global Spine J, 2011. 1(1): p. 1–8. Facchinello, Y., et al., Use of Regression Tree Analysis for Predicting the Functional Outcome after Traumatic Spinal Cord Injury . J Neurotrauma, 2021. 38(9): p. 1285–1291. Andreoli, C., et al., MRI in the acute phase of spinal cord traumatic lesions: Relationship between MRI findings and neurological outcome . Radiol Med, 2005. 110(5–6): p. 636–45. Yoo, H.J., et al., Prediction of gait recovery using machine learning algorithms in patients with spinal cord injury . Medicine (Baltimore), 2024. 103(23): p. e38286. Okimatsu, S., et al., Determining the short-term neurological prognosis for acute cervical spinal cord injury using machine learning . J Clin Neurosci, 2022. 96: p. 74–79. Wirries, A., et al., Artificial intelligence facilitates decision-making in the treatment of lumbar disc herniations . Eur Spine J, 2021. 30(8): p. 2176–2184. Kuo, B.I., et al., Keratoconus Screening Based on Deep Learning Approach of Corneal Topography . Transl Vis Sci Technol, 2020. 9(2): p. 53. You, Z., et al., Vocal cord leukoplakia classification using deep learning models in white light and narrow band imaging endoscopy images . Head Neck, 2023. 45(12): p. 3129–3145. Halligan, S., Y. Menu, and S. Mallett, Why did European Radiology reject my radiomic biomarker paper? How to correctly evaluate imaging biomarkers in a clinical setting . Eur Radiol, 2021. 31(12): p. 9361–9368. He, K.a.Z., Xiangyu and Ren, Shaoqing and Sun, Jian, Deep Residual Learning for Image Recognition. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016: p. 770–778. Ching, T., et al., Opportunities and obstacles for deep learning in biology and medicine . J R Soc Interface, 2018. 15(141). Madabhushi, A. and G. Lee, Image analysis and machine learning in digital pathology: Challenges and opportunities . Med Image Anal, 2016. 33: p. 170–175. Wang, Y., et al., The radiomic-clinical model using the SHAP method for assessing the treatment response of whole-brain radiotherapy: a multicentric study . Eur Radiol, 2022. 32(12): p. 8737–8747. Kim, Y., et al., A CT-based Deep Learning Model for Predicting Subsequent Fracture Risk in Patients with Hip Fracture . Radiology, 2024. 310(1): p. e230614. Additional Declarations No competing interests reported. Supplementary Files file.docx 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-4848654","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":348787527,"identity":"47cc1780-9a65-46c9-90e3-711677893a0a","order_by":0,"name":"Fabin Lin","email":"","orcid":"","institution":"Fujian Medical University","correspondingAuthor":false,"prefix":"","firstName":"Fabin","middleName":"","lastName":"Lin","suffix":""},{"id":348787528,"identity":"ea547441-36bb-4b06-a13b-5adf960947d2","order_by":1,"name":"Kaifeng Wang","email":"","orcid":"","institution":"Fujian Medical University","correspondingAuthor":false,"prefix":"","firstName":"Kaifeng","middleName":"","lastName":"Wang","suffix":""},{"id":348787529,"identity":"c8a59b38-526b-42da-b803-f5f422060c44","order_by":2,"name":"Ruxian Wang","email":"","orcid":"","institution":"Fujian Medical University","correspondingAuthor":false,"prefix":"","firstName":"Ruxian","middleName":"","lastName":"Wang","suffix":""},{"id":348787530,"identity":"bcfd56b2-014d-4513-9c86-07454d2c2bb2","order_by":3,"name":"Yang Wu","email":"","orcid":"","institution":"The First People’s Hospital of ChangDe City","correspondingAuthor":false,"prefix":"","firstName":"Yang","middleName":"","lastName":"Wu","suffix":""},{"id":348787531,"identity":"4dab7eff-5024-45d9-af75-343a77523972","order_by":4,"name":"Chunmei Chen","email":"","orcid":"","institution":"Fujian Medical University Union Hospital","correspondingAuthor":false,"prefix":"","firstName":"Chunmei","middleName":"","lastName":"Chen","suffix":""},{"id":348787532,"identity":"ddee2ea9-8b8f-4a8c-81e7-0cd55f62eb7d","order_by":5,"name":"Yongjiang Wang","email":"","orcid":"","institution":"Ordos Central Hosptial","correspondingAuthor":false,"prefix":"","firstName":"Yongjiang","middleName":"","lastName":"Wang","suffix":""},{"id":348787533,"identity":"8a110f24-227a-479e-88b6-27c0cf597942","order_by":6,"name":"Rui Wang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA1klEQVRIiWNgGAWjYBACeWb+779/VNjIyTMzHyBOi2E7g4E0w5k0Y8P2tgQirTkP1MLYcjix4cwZA+J0MDYzJBgXNhxmbJyR8/HGGwY7Od0GAlrYmRkOJM/ckc7MLpG72XIOQ7Kx2QGCtjA2HOA9Y83GOCN3mzQPw4HEbYS0MBxmZmzgbWPmYbiR84xYLWzMzLxtzhIMZ86wEafFsJkH6KYzaQbAQDa2nGNAhF/k+c+wMXyosKmfz8z88MabCjs5glpQgAQPkVGDrIVUHaNgFIyCUTAiAAD/SUAwW+5pBQAAAABJRU5ErkJggg==","orcid":"","institution":"Fujian Medical University Union Hospital","correspondingAuthor":true,"prefix":"","firstName":"Rui","middleName":"","lastName":"Wang","suffix":""}],"badges":[],"createdAt":"2024-08-02 13:08:53","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4848654/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4848654/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":66123548,"identity":"c49d162a-23fe-4184-8386-86973e56de2d","added_by":"auto","created_at":"2024-10-08 02:28:53","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":389757,"visible":true,"origin":"","legend":"\u003cp\u003eA. Patient selection criteria, detailing both inclusion and exclusion factors.\u003cbr\u003e\nB. Representative images of ROI segmentation examples.\u003cbr\u003e\nC. Diagram illustrating the study's methodology and progression.\u003c/p\u003e","description":"","filename":"figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-4848654/v1/966001acf4ee9514fa7317d4.png"},{"id":66123546,"identity":"61b4991f-8916-4c75-a22e-80529d914a78","added_by":"auto","created_at":"2024-10-08 02:28:52","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":206781,"visible":true,"origin":"","legend":"\u003cp\u003eROC curves for both the training and test sets. Panel A shows the ROC curve for the training set, while Panel B illustrates the ROC curve for the test set. The models are labeled as follows: RAD: Radiomics model, 18: ResNet18 model, 50: ResNet50 model, 101: ResNet101 model, RAD18: Combination of Radiomics and ResNet18, RAD50: Combination of Radiomics and ResNet50, RAD101: Combination of Radiomics and ResNet101\u003c/p\u003e","description":"","filename":"figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-4848654/v1/8baba710823ea53f2d9e5528.png"},{"id":66124605,"identity":"f82ad8a0-1cd4-425d-861b-c3428e7a87de","added_by":"auto","created_at":"2024-10-08 02:36:52","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":405187,"visible":true,"origin":"","legend":"\u003cp\u003eIn the combined ResNet + Radiomics model: (A) Decision Curve Analysis (DCA) (B) Calibration curve, comparing the predicted probabilities with the observed outcomes to assess model calibration. (C) Grad-CAM visualizations for the best-performing combined model, ResNet50 + Radiomics.\u003c/p\u003e","description":"","filename":"figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-4848654/v1/3482f898326ae228687d5a74.png"},{"id":66123544,"identity":"fede4fa3-4bb2-456b-8ef4-6c97e40e79e3","added_by":"auto","created_at":"2024-10-08 02:28:52","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":147510,"visible":true,"origin":"","legend":"\u003cp\u003eIn the Rad + ResNet50 model:\u003c/p\u003e\n\u003cp\u003eApplication of logistic regression with the Least Absolute Shrinkage and Selection Operator (LASSO) method, including the 10-fold cross-validated mean squared error (MSE) and the use of feature weights and non-zero coefficients for constructing the Rad-score. (A) Least Absolute Shrinkage and Selection Operator (LASSO)(B) MSE validated across 10 folds, demonstrating the model's consistency and performance. (C) Histogram of feature weights, showing the Rad-score derived from features selected by LASSO.\u003c/p\u003e","description":"","filename":"figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-4848654/v1/72ea71584d37b9e77cfc97a3.png"},{"id":74991107,"identity":"273296a5-01a6-41d8-8322-ad7cd719874f","added_by":"auto","created_at":"2025-01-29 07:31:58","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2623975,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4848654/v1/c032080e-7345-4201-a667-fd9660e1c50e.pdf"},{"id":66123549,"identity":"b261a883-a384-4496-b756-d614978b0ba5","added_by":"auto","created_at":"2024-10-08 02:28:53","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":3579682,"visible":true,"origin":"","legend":"","description":"","filename":"file.docx","url":"https://assets-eu.researchsquare.com/files/rs-4848654/v1/6b756f33f8d7ee2affd881be.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Postoperative One Year Prediction for Patients with Cervical Spinal Cord Injury Based on Deep Learning and Radiomics","fulltext":[{"header":"Introduction","content":"\u003cp\u003eCervical spinal cord injury (SCI) can result in a spectrum of outcomes, from complete to incomplete tetraplegia, leading to significant physical, psychological, and social consequences for those affected [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Individuals with cervical SCI experience substantial impairments in various aspects of life, requiring extensive care and effort. Preoperative prediction is crucial for postoperative patients, yet the associated risk factors remain unidentified [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe convergence of machine learning (ML) and medical advancements has significantly improved diagnosis, risk assessment, and treatment response prediction [\u003cspan additionalcitationids=\"CR5 CR6\" citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. ML algorithms can identify subtle data patterns often missed by human experts, though research on cervical SCI remains limited [\u003cspan additionalcitationids=\"CR9 CR10\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Radiomics, which extracts numerous quantitative features from medical images and analyzes them using advanced algorithms, has been employed to detect spinal cord compression at different locations and guide management decisions [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eDeep learning (DL), utilizing convolutional neural networks to extract image features, has driven significant advancements in artificial intelligence and human-computer interactions. This approach reduces diagnostic errors and maximizes the potential of radiomics data. Furthermore, deep transfer learning enables the application of deep learning to small datasets. As neuroscience advances intersect with technological innovations, neurosurgery finds unique opportunities to optimize patient care through ANNs. However, integrating clinical relevance with complex, non-linear data for accurate prognostic modeling remains challenging[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThis study aimed to compare traditional radiomics with deep learning-enhanced radiomics and to develop and validate a deep learning radiomics (DLR) model for accurately predicting cervical SCI outcomes.\u003c/p\u003e"},{"header":"Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003ePatient Selection and Study Design\u003c/h2\u003e \u003cp\u003eThis retrospective study received approval from the ethical committees of three hospitals (approval no. 2021KY138), with informed consent waived. Data from 223 patients collected between January 2012 and January 2021 were included.\u003c/p\u003e \u003cp\u003e \u003cb\u003eInclusion criteria\u003c/b\u003e:\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eConfirmed diagnosis of cervical spinal cord injury (acute injury to nerve structures in the spinal canal causing temporary or permanent sensory and motor function changes, with or without vesicorectal dysfunction)\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eComplete case data and preoperative MRI images\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eNo prior treatment before surgery\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eExclusion criteria\u003c/strong\u003e \u003cp\u003eIncomplete clinical records or inadequate imaging data.\u003c/p\u003e \u003c/p\u003e \u003cp\u003eInitially, 223 patients were included: 118 from Ordos Central Hospital, and 19 from The First People\u0026rsquo;s Hospital of ChangDe City, 86 from Fujian Medical University Union Hospital. We excluded 86 patients with non-cervical spinal cord injuries. Of the remaining 137 patients, 55 were excluded due to incomplete or low-quality MRI images. Ultimately, 82 patients were analyzed, divided into good prognosis (41 patients) and poor prognosis (41 patients). The patient recruitment map is shown in Fig.\u0026nbsp;1A, and the study design and pipeline are illustrated in Fig.\u0026nbsp;1C.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eSurgical Procedures\u003c/h2\u003e \u003cp\u003e All participants underwent neck surgery via either an anterior or posterior approach, including anterior cervical discectomy, microdecompression of the spinal canal, implant fusion, internal fixation with steel plate, or posterior median cervical internal fixation, implant fusion, and decompression. Based on the study by Yanlin Yin et al., there were no significant differences in postoperative cervical and neurological recovery between anterior and posterior surgical interventions[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Follow-ups were conducted to monitor progress. Detailed descriptions of the surgical techniques are provided below:\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003ePatient Position\u003c/b\u003e: Prone for the posterior approach, supine for the anterior approach.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eAnesthesia Method\u003c/b\u003e: Tracheal intubation with sedation combined with general anesthesia.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003ePreoperative Localization\u003c/b\u003e: Head frame fixation after retraction (posterior approach) and immobilization. Using X-ray front and side fluoroscopy of the C-arm machine, the surgical segments were localized and marked.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eLocalization of Puncture\u003c/b\u003e: After routine disinfection and towel spreading, the surgical incision was made according to intraoperative C-arm X-ray positive and lateral fluoroscopy film, with layer-by-layer incision of the skin, subcutaneous tissue, and deep fascia.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eSurgical Steps\u003c/b\u003e[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]:\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eAnterior Approach\u003c/b\u003e: Vertebral body resection and thorough excision of implicated intervertebral discs. Cervical fusion was achieved using either allograft fibular struts or autografts, secured with semi-rigid locking plates. Postoperatively, cervical immobilization with a collar was maintained for two weeks until wound healing.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003ePosterior Approach\u003c/b\u003e: Using a Mayfield clamp or horseshoe headrest, the cervical spine was stabilized in a neutral position. The extent of cervical laminoplasty was predetermined through preoperative imaging. Transverse mass screws were deployed and affixed with connecting rods to achieve segmental decompression by excising the ligamenta flava and laminae of the targeted segments.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003e\u003col\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003ePostoperative Care\u003c/b\u003e: Complete hemostasis was ensured, paravertebral muscles were repositioned, and the subcutaneous layer and skin were sutured. The operative site was dressed with sterile gauze, and a drainage tube was inserted. Patients were encouraged to start range-of-motion exercises immediately after the drainage tube was removed and were discharged once they were able to walk. Both cohorts were managed in alignment with the hospital\u0026rsquo;s clinical guidelines for cervical spine interventions, incorporating a uniform regimen of physical therapy.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003c/ol\u003e\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eClinical Data Collection and ASIA Scale Assessment\u003c/h2\u003e \u003cp\u003ePatient demographic and clinical information, including age, gender, and cause of spinal cord injury, were collected. The preoperative and postoperative ASIA scales were utilized to calculate the difference. During postoperative follow-up, the ASIA score criteria remained consistent with preoperative standards. Patients were classified into good prognosis and poor prognosis groups based on whether their ASIA grade improved by at least one level post-surgery: a decrease indicated a good prognosis, otherwise poor prognosis. Regarding the ASIA grade scoring criteria, please refer to Supplementary File 1.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eAcquisition and Preprocessing of T2-weighted MRI Images\u003c/h2\u003e \u003cp\u003eThe preoperative MRI scans for all enrolled patients, featuring T2-weighted imaging (T2WI), were conducted using 3.0 Tesla MRI technology. The machine settings at different centers are shown in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e3\u003c/span\u003e. After obtaining the T2-weighted MRI images, preprocessing included normalization to the range of 0\u0026ndash;1, resampling to a voxel size of 1x1x1, and processing image noise using multiple filters: Median, Average Gaussian and Box filtering. Each filter outputted as a separate channel rather than being overlaid. These operations aimed at reducing multicenter effects and enhancing the model's generalization ability. Detailed descriptions of each filter can be found in the supplementary file2.\u003c/p\u003e\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eROI Segmentation and Radiomics Analysis\u003c/h2\u003e \u003cdiv id=\"Sec8\" class=\"Section3\"\u003e \u003ch2\u003eSegmentation of Regions of Interest (ROI)\u003c/h2\u003e \u003cp\u003eThe 3D Slicer Software Application v5.0.2 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.slicer.org/\u003c/span\u003e\u003cspan address=\"https://www.slicer.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e)\u003c/span\u003e was used to segment the Region of Interest (ROI). The ROI comprised two segments: the spinal cord injury region (segmentation 1) and the unaffected spinal cord regions (segmentation 2). Segmentation 1 is defined as the area of evident spinal cord injury, such as edema and significant spinal cord compression [\u003cspan additionalcitationids=\"CR16\" citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. Segmentation 2 is defined as the region excluding the anatomical landmarks of Segmentation 1, including the upper end level aligned with the foramen magnum and the lower end level aligned with the lower edge of the first lumbar vertebra in adults. This defined the extent of the ROI, encompassing the entire spinal cord while carefully excluding vertebral and cerebrospinal fluid regions. This approach ensures that significant biomarkers outside the area of spinal cord injury are not overlooked. Figure\u0026nbsp;1B illustrates our ROI delineation. Consistency between original images and the four filtering techniques ensured uniformity in ROI delineation for each patient.\u003c/p\u003e \u003cp\u003eEach participant's ROIs were independently segmented by two raters, R.W. and Y.W. Intra-rater reliability was assessed by the same radiologist segmenting 30 cases two weeks apart, while inter-rater reliability was evaluated by another radiologist segmenting the same 30 cases. Intraclass correlation coefficients (ICCs) were calculated to assess both intra- and inter-rater reliability, with an ICC\u0026thinsp;\u0026gt;\u0026thinsp;0.75 indicating strong consistency.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003eExtraction of Radiomics Features\u003c/h2\u003e \u003cp\u003eUsing the Pyradiomics module within 3D Slicer, radiomics features were extracted and organized into seven categories for each segmentation in each filtering technique: (1) Shape-based features: 14, (2) First-order features: 18, (3) Gray-level dependence matrix (GLDM) features: 14, (4) Gray-level size zone matrix (GLSZM) features: 16, (5) Neighboring gray-tone difference matrix (NGTDM) features: 5, (6) Gray-level run-length matrix (GLRLM) features: 16, and (7) Gray-level co-occurrence matrix (GLCM) features: 24. Processing with the four filtering techniques yielded a total of 1,070 radiomics features per patient. Detailed feature categories are provided in Supplementary File 3 and4.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eDeep learning analysis\u003c/h2\u003e \u003cdiv id=\"Sec11\" class=\"Section3\"\u003e \u003ch2\u003eTraining Deep Learning Models\u003c/h2\u003e \u003cp\u003eFor training the deep learning models, the sagittal plane image with the maximum ROI for each patient was selected and resampled to 224x224 pixels. Other image data preparation steps followed the same preprocessing process as described above. Three deep learning models were used: ResNet18, ResNet50, and ResNet101, all pretrained on the ImageNet dataset [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. Five different channels were input into each model: the original image, box, average, Gaussian and median filtering. Each channel underwent an independent yet identical training process with consistent parameter settings. The five separately trained channels ultimately converged to contribute to the classification task. Data were split into training and test sets in a 7:3 ratio. We have carefully set the parameters (initial learning rate: 0.01, batch: 64, 50 epochs). The classification targets were good prognosis (1) and poor prognosis (0).\u003c/p\u003e \u003cp\u003eModel parameters were divided into two parts: 1. the backbone layer (backbone) and 2. the task-specific layer (task-spec). Task-specific parameters were randomly initialized, while backbone parameters utilized the pre-trained model parameters from ImageNet. For the task-specific parameters, the cosine annealing learning rate decay algorithm was applied [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e], with details provided in the supplementary file5.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eDeep learning feature extraction\u003c/h2\u003e \u003cp\u003eThe penultimate layer of the model (AveragePooling layer) was used for feature extraction. Our dataset, which includes five categories of images, was independently trained through five identical deep learning channels. Deep learning features were extracted separately for each channel. For ResNet50 and ResNet101 the feature dimensionality was 2048, while for and ResNet18, it was 512. To reduce the risk of overfitting and improve the model's generalization capability, principal component analysis (PCA) was applied to compress the extracted features to 32 dimensions per channel, resulting in 160 deep learning features (32x5) per patient.\u003c/p\u003e \u003cp\u003eIn multi-center studies, the ComBat compensation procedure addresses differences arising from imaging protocols and various MRI scanners[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. This method shows considerable promise in improving the reproducibility of research conducted across multiple centers.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eFeature engineering\u003c/h2\u003e \u003cp\u003eFirst, We utilized the ComBat compensation procedure on all extracted features to mitigate multi-center effects[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. After that, Radiomics features (1,070) were combined with deep learning features (160), resulting in 1,230 fused features per patient. Before feature selection, data were standardized using Z-scores to convert all features to a uniform mean of 0 and a standard deviation of 1.\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eInitial Screening\u003c/b\u003e: Features with a p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 were retained using the Mann-Whitney U test.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eRedundancy Reduction\u003c/b\u003e: Features were screened for high repeatability using the Spearman rank correlation coefficient, retaining only one feature from any pair with a correlation coefficient greater than 0.9[\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. A greedy recursive deletion strategy was applied to remove the feature with the highest redundancy in each iteration.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eFeature Selection\u003c/b\u003e: The Minimum Redundancy Maximum Relevance (MRMR) algorithm was used to select the best features. A 10-fold cross-validated Least Absolute Shrinkage and Selection Operator (LASSO) regression identified features with non-zero coefficients for subsequent modeling.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eDevelopment and assessment of models\u003c/h2\u003e \u003cp\u003eThe study employed MLP machine learning methods, implemented using the scikit-learn package (version: 0.18) in Python 3.9. The dataset was randomly split into training (70%) and test (30%) datasets to evaluate model accuracy. Five-fold cross-validation and Grid Search were used to determine the optimal hyperparameters for each model. Evaluation metrics included AUC, Decision Curve Analysis (DCA), Calibration curve, specificity, sensitivity, and accuracy, among others.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eCategorical variables and normally distributed variables were analyzed using the chi-square test and t-tests, respectively. A p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 indicated statistical significance. Python (version 3.9, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.python.org\u003c/span\u003e\u003cspan address=\"http://www.python.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e)\u003c/span\u003e was utilized for all analyses.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003ePatient characteristics\u003c/h2\u003e \u003cp\u003eA total of 82 patients Cervical SCI were enrolled in this study, who were aged from 9 to 84 years (a mean age of 52 years). As shown in the methods section. The entire dataset was divided into two distinct groups: good prognosis and bad prognosis. Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e1\u003c/span\u003e summarizes the comparison of demographic characteristics of the two groups.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eClinical characteristics\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFeature_name\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDifference in ASI number, no improvement(n\u0026thinsp;=\u0026thinsp;41)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eDifference in ASI number, improvement(n\u0026thinsp;=\u0026thinsp;41)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003epvalue\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e52.33\u0026thinsp;\u0026plusmn;\u0026thinsp;12.32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e52.83\u0026thinsp;\u0026plusmn;\u0026thinsp;11.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e51.83\u0026thinsp;\u0026plusmn;\u0026thinsp;13.63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.593625\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGender\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.567806\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e15(18.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e9(21.95)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e6(14.63)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e67(81.71)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e32(78.05)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e35(85.37)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInjury mechanism:\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.082036\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFall on the same level\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e18(21.95)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13(31.71)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5(12.20)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFall from a height less than 1 meter\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2(2.44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1(2.44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1(2.44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFall from a height greater than 1 meter\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e37(45.12)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e14(34.15)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e23(56.10)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTraffic accident\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e17(20.73)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e7(17.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e10(24.39)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOthers (e.g., sports)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8(9.76)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6(14.63)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2(4.88)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eNumber of injured segments\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.452569\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e22(26.83)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13(31.71)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e9(21.95)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e36(43.90)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e16(39.02)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e20(48.78)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e14(17.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6(14.63)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e8(19.51)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7(8.54)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5(12.20)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2(4.88)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1(1.22)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1(2.44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1(1.22)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1(2.44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1(1.22)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1(2.44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFracture\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.03609\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e54(65.85)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e32(78.05)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e22(53.66)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e28(34.15)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e9(21.95)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e19(46.34)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePreoperative_ASIA_grade\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrade A\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e13(15.48)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6(14.63)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e7(16.28)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrade B\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e15(17.86)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4(9.76)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e11(25.58)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrade C\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e24(28.57)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4(9.76)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e20(46.51)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrade D\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e26(30.95)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e21(51.22)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5(11.63)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrade E\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6(7.14)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6(14.63)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePostoperative_ASIA_grade\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.0275\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrade A\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7(8.33)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e7(17.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrade B\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7(8.33)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3(7.32)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4(9.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrade C\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12(14.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e8(19.51)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4(9.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrade D\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e48(57.14)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e19(46.34)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e29(67.44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrade E\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e10(11.90)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4(9.76)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e6(13.95)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eForecasting with the Deep earning and radiomics model\u003c/h2\u003e \u003cdiv id=\"Sec19\" class=\"Section3\"\u003e \u003ch2\u003eRadiomics features and modeling\u003c/h2\u003e \u003cp\u003eA total of 1045 features passed the ICC test. After extensive feature screening, seven features were incorporated into the radiomics model. The model demonstrated an AUC of 0.869 (95% CI 0.7862\u0026ndash;0.9525) in the training cohort, with sensitivity (0.875) and specificity (0.727), respectively. In the test cohort, the AUC was 0.7780 (95% CI 0.5285-1), with sensitivity (0.778) and specificity (0.750). Detailed statistical results and curves are presented in Fig.\u0026nbsp;2 and Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eComparision results of performance different Deep learning models\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"12\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c11\" colnum=\"11\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c12\" colnum=\"12\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eModel\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCohort\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAUC\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e95% CI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eAccuracy\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eSen\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eSpe\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003ePPV\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eNPV\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003ePre\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c11\"\u003e \u003cp\u003eRecall\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c12\"\u003e \u003cp\u003eF1\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eRad\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003etrain\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.8690\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.7862\u0026ndash;0.9525\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.8000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.7270\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.7570\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.8570\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.7570\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c12\"\u003e \u003cp\u003e0.8120\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003etest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.7780\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.5285\u0026ndash;1.0000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.7650\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.7780\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.7500\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.7780\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.7500\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.7780\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.7780\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c12\"\u003e \u003cp\u003e0.7780\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eResNet18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003etrain\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.9480\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.9014\u0026ndash;0.9945\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.8770\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.8480\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.9060\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.9030\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.8530\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.9030\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.8480\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c12\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003etest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.7920\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.5695\u0026ndash;1.0000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.7650\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.6670\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.7000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.8570\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.7000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c12\"\u003e \u003cp\u003e0.7780\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eRad+\u003c/p\u003e \u003cp\u003eResNet18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003etrain\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.8570\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.7703\u0026ndash;0.9437\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.7690\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.6250\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.9090\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.8700\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.7140\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.8700\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.6250\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c12\"\u003e \u003cp\u003e0.7270\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003etest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.8470\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.6474\u0026ndash;1.0000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.8240\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.7780\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.7780\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.7780\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c12\"\u003e \u003cp\u003e0.8240\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eResNet50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003etrain\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.9750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.9358\u0026ndash;1.0000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.9550\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.9390\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.9700\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.9690\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.9410\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.9690\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.9390\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c12\"\u003e \u003cp\u003e0.9540\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003etest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.6898\u0026ndash;1.0000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c12\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eRad+\u003c/p\u003e \u003cp\u003eResNet50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003etrain\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.9400\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.8835\u0026ndash;0.9971\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.9240\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.9090\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.9390\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.9370\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.9120\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.9370\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.9090\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c12\"\u003e \u003cp\u003e0.9230\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003etest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.9220\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.7919\u0026ndash;1.0000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c12\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eResNet101\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003etrain\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.9270\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.8682\u0026ndash;0.9849\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.8640\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.8790\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.8480\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.8530\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.8530\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.8790\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c12\"\u003e \u003cp\u003e0.8660\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003etest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.8590\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.6692\u0026ndash;1.0000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.8120\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.7500\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.7780\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.8570\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.7780\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.8750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c12\"\u003e \u003cp\u003e0.8240\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eRad+\u003c/p\u003e \u003cp\u003eResNet101\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003etrain\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.8820\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.8030\u0026ndash;0.9602\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.8150\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.8790\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.7500\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.7840\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.8570\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.7840\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.8790\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c12\"\u003e \u003cp\u003e0.8290\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \n \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eMRI Scanners in Three Centers\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eItem\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFujian Medical University Union Hospital\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOrdos Central Hospital\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eThe First People\u0026rsquo;s Hospital of ChangDe City\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMRI Machine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGE Signa\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGE Signa\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eGE Signa\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTR Settings (ms)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2500\u0026ndash;3000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2000\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTE Settings (ms)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e100\u0026ndash;110\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e120\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e80\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSlice Thickness (mm)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSlice Spacing (mm)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMatrix Configuration\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e256 \u0026times; 512\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e256 \u0026times; 512\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e256 \u0026times; 512\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eField of View (FOV) (mm \u0026times; mm)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e240 \u0026times; 240\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e240 \u0026times; 240\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e240 \u0026times; 240\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eImaging Plane\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSagittal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSagittal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSagittal\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \n\u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eDeep learning model\u003c/h2\u003e \u003cp\u003eThe features used to build ResNet-18, ResNet-50, and ResNet-101 were 10, 9, and 10, respectively. With regard to the predictions generated by the deep learning models, among the three models we utilized, ResNet50 performed the best, followed by ResNet101. In the test set, the AUCs for ResNet50, ResNet101, and ResNet18 were 0.8750, 0.8590, and 0.7920, respectively; their accuracies were 0.8750, 0.8120, and 0.7650; sensitivities were 0.8750, 0.8750, and 0.8750; specificities were 0.8750, 0.7500, and 0.6670. (Fig.\u0026nbsp;2, Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Supplementary File 6 illustrates the Lasso regression paths and Mean Squared Error (MSE) plots for each model. Supplementary File 7 provides the feature weights utilized in constructing these models.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003eCombined model\u003c/h2\u003e \u003cp\u003eConsidering the number of features used to establish each model, the radiomics\u0026thinsp;+\u0026thinsp;ResNet18 and radiomics\u0026thinsp;+\u0026thinsp;ResNet50 models embraced the inclusion of nine features(Figure4), while the radiomics\u0026thinsp;+\u0026thinsp;ResNet101 model was constructed using eight select features.\u003c/p\u003e \u003cp\u003eAmong these, the Rad\u0026thinsp;+\u0026thinsp;ResNet50 model also performed the best: In the test set, the AUCs for Rad\u0026thinsp;+\u0026thinsp;ResNet50, Rad\u0026thinsp;+\u0026thinsp;ResNet101, and Rad\u0026thinsp;+\u0026thinsp;ResNet18 were 0.9220, 0.8750, and 0.8470, respectively; their accuracies were 0.8750, 0.8820, and 0.8240; sensitivities were 0.8750, 0.8750, and 0.7780; specificities were 0.8750, 0.8890, and 0.8750(Fig.\u0026nbsp;2). For more details, please reference Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e \u003cp\u003eDCA provided additional supporting evidence for these conclusions. DCA shows that when the threshold probability surpasses the 20% mark, the model produces more net benefit in the current study. The Decision Curve Analysis (DCA) and calibration plots indicate strong clinical utility. (Fig.\u0026nbsp;3A-B)\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003eGrad-CAM\u003c/h2\u003e \u003cp\u003eVisualizing the final convolutional layer using Grad-CAM is an effective method to enhance model interpretability. As illustrated in Fig.\u0026nbsp;3C, the darker blue regions highlight crucial areas for prognosis prediction. These regions are primarily focused on the edematous spinal cord, deformed vertebrae, and swollen ligaments.\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, we included 82 patients with cervical spinal cord injury (SCI). We discovered that the predictive capability of the feature fusion model was enhanced compared to using radiomics or CNNs individually. Ultimately, we found that the Rad\u0026thinsp;+\u0026thinsp;ResNet-50 model had the highest prognostic predictive value for cervical SCI, with AUCs of 0.940 and 0.922 in the training and test cohorts, respectively.\u003c/p\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003eContributions of This Study and Comparison with Previous Studies\u003c/h2\u003e \u003cp\u003eFunctional prognosis after cervical spinal cord injury (SCI), particularly regarding regaining functional independence, is a primary focus in rehabilitation. Predicting SCI recovery can be challenging, but the ASIA grade is a key benchmark for assessing clinical recovery and long-term prognosis [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. In our study, 82 cervical SCI patients were divided into two groups based on whether their ASIA index improved after one year. Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e1\u003c/span\u003e shows preoperative and postoperative ASIA grades with p\u0026thinsp;\u0026lt;\u0026thinsp;0.05, indicating clinical significance. Yann Facchinello et al. also highlight that the severity of the neurological deficit at admission, indicated by the ASIA grade, is a critical predictor of recovery [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eIn addition to the ASIA score, MRI images are crucial for prognosis, with metrics shown to be reproducible and predictive of clinical outcomes[\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e].Hyun-Joon Yoo et al. used machine learning algorithms to create a prediction model for SCI patients based solely on clinical variables, achieving promising results [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. However, their approach did not include the prognostic value of imaging, particularly MRI.\u003c/p\u003e \u003cp\u003eThe BASIC (Brain and Spinal Injury Center) score, developed to predict neurological improvement based on high signal intensity on axial T2-weighted MRI, has proven effective [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e] [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. However, this score focused on a limited set of manually measured MRI features, such as injury length and spinal canal compromise, overlooking many other potential features.\u003c/p\u003e \u003cp\u003eOur study leverages deep learning radiomics to extract and model MRI features. This approach reduces the risk of missing significant features, as evidenced by the combined model (Rad\u0026thinsp;+\u0026thinsp;ResNet50 AUC: 0.9220), which demonstrates improved performance.\u003c/p\u003e \u003cp\u003eWe noted that in another study using deep learning, Okimatsu and colleagues developed a CNN model to quantify radiographic characteristics and predict 1-month neurological outcomes in acute SCI patients, achieving 71% accuracy[\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. We extended the follow-up period (1-year) and also improved the model's accuracy to 92.2%. Integrating handcrafted radiomics (HCR) and deep transfer learning (DTL) features for cervical SCI prognosis is novel, and our approach uses standard image data without special training, highlighting its potential.\u003c/p\u003e \u003cp\u003eTraining deep learning models on small datasets has limitations, though deep transfer learning can help. Andr\u0026eacute; Wirries suggested that using small datasets to train deep learning models for clinical applications is practical[\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. We enhanced model performance through random shifts, rotations, and horizontal flips during training to increase data diversity[\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. Expanding the dataset and controlling feature numbers further improved results. We implemented rigorous feature selection processes in our experiments[\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eResNet\u0026rsquo;s structure uses shortcut connections, enabling effective fusion training[\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. To evaluate different models, we tested ResNet-18, ResNet-50, and ResNet-101. Both deep learning and combined models showed that ResNet-50 performed best. Despite having fewer layers and higher Top-1 and Top-5 error rates than ResNet-101[\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e], ResNet-50\u0026rsquo;s moderate depth proved advantageous in this research. This indicates that while deeper networks may offer superior learning capabilities, selecting an appropriate network depth for specific research can lead to better model fitting and other benefits, such as reduced computational resources and time. Additionally, ResNet-18 underperformed compared to ResNet-50, suggesting lower-dimensional ResNet networks may lack sufficient fitting capacity.\u003c/p\u003e \u003cdiv id=\"Sec25\" class=\"Section3\"\u003e \u003ch2\u003eDeep Learning Model Interpretability\u003c/h2\u003e \u003cp\u003eAlthough deep learning shows promise in disease diagnosis and prognosis, the interpretability of these models can hinder their broader application[\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. Some studies use post-hoc methods or supervised machine learning models to interpret deep learning algorithm outputs[\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. For example, Yixin Wang visualized SVM features using a SHAP-based method[\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. In our study, we generated Grad-CAM images to enhance model interpretability, similar to how Kim, Y. and colleagues used heatmaps to identify regions crucial for prognosis prediction[\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec26\" class=\"Section3\"\u003e \u003ch2\u003elimitations\u003c/h2\u003e \u003cp\u003eThere are some limitations to this study that further studied and explored in future work. First, the sample size of cervical SCI was small. Although our findings reflect the predictive ability of deep learning features to a certain extent, more data would be more convincing. Second, this was a restrospective study, more prospective data are needed to verify the effectiveness of the model.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Conclusions","content":"\u003cp\u003eIn conclusion, the DLR features were fused with cervical SCI features based on MRI images, which improved the identification ability of a single radiomics prognostic model for cervical SCI. And we developed a new prognostic model (Rad\u0026thinsp;+\u0026thinsp;ResNet-50) which had the highest predictive value for prognosis of cervical SCI with AUCs of 0.940 and 0.922 in the training cohort and test cohort, respectively.\u003c/p\u003e "},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eFunding/Support:\u003c/strong\u003e This research received no external funding\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of Interest:\u003c/strong\u003e The authors have nothing to disclose.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResearch Ethics:\u003c/strong\u003e Institutional Review Board approval was obtained. The study was conducted according to the guidelines of the Declaration of Helsinki, and was approved by Institutional Review Board (approval no. 2021KY138).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgments:\u003c/strong\u003e We thank the colleagues in our department for their help in our study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eInformed Consent Statement:\u0026nbsp;\u003c/strong\u003eWritten informed consent was not required for this study because this is a retrospective study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e：The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAlizadeh, A., S.M. Dyck, and S. Karimi-Abdolrezaee, \u003cem\u003eTraumatic Spinal Cord Injury: An Overview of Pathophysiology, Models and Acute Injury Mechanisms\u003c/em\u003e. Front Neurol, 2019. 10: p. 282.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAarabi, B., et al., \u003cem\u003eIntramedullary Lesion Length on Postoperative Magnetic Resonance Imaging is a Strong Predictor of ASIA Impairment Scale Grade Conversion Following Decompressive Surgery in Cervical Spinal Cord Injury\u003c/em\u003e. Neurosurgery, 2017. 80(4): p. 610\u0026ndash;620.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWichmann, T.O., et al., \u003cem\u003eEarly clinical predictors of functional recovery following traumatic spinal cord injury: a population-based study of 143 patients\u003c/em\u003e. Acta Neurochirurgica, 2021. 163(8): p. 2289\u0026ndash;2296.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGiger, M.L., \u003cem\u003eMachine Learning in Medical Imaging.\u003c/em\u003e J Am Coll Radiol, 2018. 15(3 Pt B): p. 512\u0026ndash;520.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWong, K.K.L., L. Wang, and D. Wang, \u003cem\u003eRecent developments in machine learning for medical imaging applications\u003c/em\u003e. Comput Med Imaging Graph, 2017. 57: p. 1\u0026ndash;3.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEnglish, M., et al., \u003cem\u003eMachine Learning in Neuro-Oncology, Epilepsy, Alzheimer's Disease, and Schizophrenia\u003c/em\u003e. Acta Neurochir Suppl, 2022. 134: p. 349\u0026ndash;361.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDietz, N., et al., \u003cem\u003eEvaluation of Predictive Models for Complications following Spinal Surgery\u003c/em\u003e. J Neurol Surg A Cent Eur Neurosurg, 2020. 81(6): p. 535\u0026ndash;545.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eInoue, T., et al., \u003cem\u003eXGBoost, a Machine Learning Method, Predicts Neurological Recovery in Patients with Cervical Spinal Cord Injury\u003c/em\u003e. Neurotrauma Rep, 2020. 1(1): p. 8\u0026ndash;16.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDietz, N., et al., \u003cem\u003eMachine learning in clinical diagnosis, prognostication, and management of acute traumatic spinal cord injury (SCI): A systematic review\u003c/em\u003e. J Clin Orthop Trauma, 2022. 35: p. 102046.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHaefeli, J., et al., \u003cem\u003eMultivariate Analysis of MRI Biomarkers for Predicting Neurologic Impairment in Cervical Spinal Cord Injury\u003c/em\u003e. AJNR Am J Neuroradiol, 2017. 38(3): p. 648\u0026ndash;655.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhan, O., et al., \u003cem\u003eUse of Machine Learning and Artificial Intelligence to Drive Personalized Medicine Approaches for Spine Care\u003c/em\u003e. World Neurosurg, 2020. 140: p. 512\u0026ndash;518.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAzimi, P., et al., \u003cem\u003eA Review on the Use of Artificial Intelligence in Spinal Diseases\u003c/em\u003e. Asian Spine J, 2020. 14(4): p. 543\u0026ndash;571.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYin, Y., et al., \u003cem\u003eThe Efficacy of Anterior Cervical Corpectomy and Fusion and Posterior Total Laminectomy on Cervical Spinal Cord Injury and Quality of Life.\u003c/em\u003e Comput Math Methods Med, 2022. 2022: p. 8216339.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSeng, C., et al., \u003cem\u003eSurgically treated cervical myelopathy: a functional outcome comparison study between multilevel anterior cervical decompression fusion with instrumentation and posterior laminoplasty\u003c/em\u003e. Spine J, 2013. 13(7): p. 723\u0026ndash;31.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTarawneh, A.M., et al., \u003cem\u003eCan MRI findings predict the outcome of cervical spinal cord Injury? a systematic review\u003c/em\u003e. European Spine Journal, 2020. 29(10): p. 2457\u0026ndash;2464.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, M.Z., et al., \u003cem\u003ePredicting postoperative recovery in cervical spondylotic myelopathy: construction and interpretation of T2(*)-weighted radiomic-based extra trees models\u003c/em\u003e. Eur Radiol, 2022. 32(5): p. 3565\u0026ndash;3575.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBurns, J.E., J. Yao, and R.M. Summers, \u003cem\u003eVertebral Body Compression Fractures and Bone Density: Automated Detection and Classification on CT Images\u003c/em\u003e. Radiology, 2017. 284(3): p. 788\u0026ndash;797.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOrlhac, F., et al., \u003cem\u003eValidation of A Method to Compensate Multicenter Effects Affecting CT Radiomics\u003c/em\u003e. Radiology, 2019. 291(1): p. 53\u0026ndash;59.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLucia, F., et al., \u003cem\u003eExternal validation of a combined PET and MRI radiomics model for prediction of recurrence in cervical cancer patients treated with chemoradiotherapy\u003c/em\u003e. Eur J Nucl Med Mol Imaging, 2019. 46(4): p. 864\u0026ndash;877.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBi, S., et al., \u003cem\u003eMulti-parametric MRI-based radiomics signature for preoperative prediction of Ki-67 proliferation status in sinonasal malignancies: a two-centre study\u003c/em\u003e. Eur Radiol, 2022.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, W., et al., \u003cem\u003eDevelopment and Validation of a Computed Tomography-Based Radiomics Signature to Predict Response to Neoadjuvant Chemotherapy for Locally Advanced Gastric Cancer\u003c/em\u003e. JAMA Netw Open, 2021. 4(8): p. e2121143.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Middendorp, J.J., et al., \u003cem\u003eDiagnosis and prognosis of traumatic spinal cord injury\u003c/em\u003e. Global Spine J, 2011. 1(1): p. 1\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFacchinello, Y., et al., \u003cem\u003eUse of Regression Tree Analysis for Predicting the Functional Outcome after Traumatic Spinal Cord Injury\u003c/em\u003e. J Neurotrauma, 2021. 38(9): p. 1285\u0026ndash;1291.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndreoli, C., et al., \u003cem\u003eMRI in the acute phase of spinal cord traumatic lesions: Relationship between MRI findings and neurological outcome\u003c/em\u003e. Radiol Med, 2005. 110(5\u0026ndash;6): p. 636\u0026ndash;45.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYoo, H.J., et al., \u003cem\u003ePrediction of gait recovery using machine learning algorithms in patients with spinal cord injury\u003c/em\u003e. Medicine (Baltimore), 2024. 103(23): p. e38286.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOkimatsu, S., et al., \u003cem\u003eDetermining the short-term neurological prognosis for acute cervical spinal cord injury using machine learning\u003c/em\u003e. J Clin Neurosci, 2022. 96: p. 74\u0026ndash;79.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWirries, A., et al., \u003cem\u003eArtificial intelligence facilitates decision-making in the treatment of lumbar disc herniations\u003c/em\u003e. Eur Spine J, 2021. 30(8): p. 2176\u0026ndash;2184.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKuo, B.I., et al., \u003cem\u003eKeratoconus Screening Based on Deep Learning Approach of Corneal Topography\u003c/em\u003e. Transl Vis Sci Technol, 2020. 9(2): p. 53.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYou, Z., et al., \u003cem\u003eVocal cord leukoplakia classification using deep learning models in white light and narrow band imaging endoscopy images\u003c/em\u003e. Head Neck, 2023. 45(12): p. 3129\u0026ndash;3145.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHalligan, S., Y. Menu, and S. Mallett, \u003cem\u003eWhy did European Radiology reject my radiomic biomarker paper? How to correctly evaluate imaging biomarkers in a clinical setting\u003c/em\u003e. Eur Radiol, 2021. 31(12): p. 9361\u0026ndash;9368.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHe, K.a.Z., Xiangyu and Ren, Shaoqing and Sun, Jian, \u003cem\u003eDeep Residual Learning for Image Recognition.\u003c/em\u003e 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016: p. 770\u0026ndash;778.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChing, T., et al., \u003cem\u003eOpportunities and obstacles for deep learning in biology and medicine\u003c/em\u003e. J R Soc Interface, 2018. 15(141).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMadabhushi, A. and G. Lee, \u003cem\u003eImage analysis and machine learning in digital pathology: Challenges and opportunities\u003c/em\u003e. Med Image Anal, 2016. 33: p. 170\u0026ndash;175.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, Y., et al., \u003cem\u003eThe radiomic-clinical model using the SHAP method for assessing the treatment response of whole-brain radiotherapy: a multicentric study\u003c/em\u003e. Eur Radiol, 2022. 32(12): p. 8737\u0026ndash;8747.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim, Y., et al., \u003cem\u003eA CT-based Deep Learning Model for Predicting Subsequent Fracture Risk in Patients with Hip Fracture\u003c/em\u003e. Radiology, 2024. 310(1): p. e230614.\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":"","lastPublishedDoi":"10.21203/rs.3.rs-4848654/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4848654/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground:\u003c/h2\u003e \u003cp\u003eCervical spinal cord injury (SCI) can lead to significant impairments, requiring extensive care and posing considerable challenges in predicting postoperative outcomes. This study aimed to develop and validate a deep learning radiomics (DLR) model combining deep learning and radiomics features to improve the prognostic prediction of cervical SCI.\u003c/p\u003e\u003ch2\u003eMethods:\u003c/h2\u003e \u003cp\u003eThis retrospective study included 82 patients with confirmed cervical SCI from three hospitals, collected between January 2012 and January 2021. Patients were divided into good prognosis and poor prognosis groups based on postoperative ASIA grade improvement. Preoperative MRI images were processed using various filtering techniques, and regions of interest (ROI) were segmented and analyzed to extract radiomics features. Deep learning models (ResNet-18, ResNet-50, and ResNet-101) were trained. Features from both radiomics and deep learning models were combined and selected 、 to build the final predictive model using MLP.\u003c/p\u003e\u003ch2\u003eResults:\u003c/h2\u003e \u003cp\u003eResNet-50 outperformed other models, demonstrating an AUC of 0.8750 in the test set. The combined model (Rad\u0026thinsp;+\u0026thinsp;ResNet-50) showed the highest prognostic value with an AUC of 0.9220 in the test set. Grad-CAM images enhanced the interpretability of the model by highlighting critical areas for prognosis prediction.\u003c/p\u003e\u003ch2\u003eConclusion:\u003c/h2\u003e \u003cp\u003eIntegrating deep learning and radiomics features significantly improves the prediction accuracy for cervical SCI outcomes. The Rad\u0026thinsp;+\u0026thinsp;ResNet-50 model, with its superior performance and interpretability, holds promise for clinical applications, offering a robust tool for predicting functional prognosis in cervical SCI patients. Further prospective studies with larger datasets are needed to validate these findings.\u003c/p\u003e","manuscriptTitle":"Postoperative One Year Prediction for Patients with Cervical Spinal Cord Injury Based on Deep Learning and Radiomics","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-10-08 02:28:48","doi":"10.21203/rs.3.rs-4848654/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":"a7085c28-f895-4c82-9baa-8340b1682282","owner":[],"postedDate":"October 8th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":37009719,"name":"Health sciences/Neurology/Neurological disorders/Neuromuscular disease"},{"id":37009720,"name":"Biological sciences/Neuroscience/Diseases of the nervous system/Neurodegeneration"}],"tags":[],"updatedAt":"2025-01-29T07:23:46+00:00","versionOfRecord":[],"versionCreatedAt":"2024-10-08 02:28:48","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4848654","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4848654","identity":"rs-4848654","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}