An interpretable multimodal model integrating clinical, spectral CT imaging, and deep learning analysis of intra- and peritumoral regions for preoperative prediction of perineural invasion in gastric cancer: A prospective, multicenter study.
Authors
Affiliations (12)
Affiliations (12)
- Department of Radiology, Fujian Medical University Union Hospital, Fuzhou 350001, China; The School of Medical Imaging, Fujian Medical University, Fuzhou 350122, China.
- Department of Radiology, Fujian Medical University Union Hospital, Fuzhou 350001, China; The School of Medical Imaging, Fujian Medical University, Fuzhou 350122, China; Fujian Key Laboratory of Intelligent Imaging and Precision Radiotherapy for Tumors (Fujian Medical University), Fuzhou 350001, China.
- Department of Radiology, Fujian Medical University Union Hospital, Fuzhou 350001, China; The School of Medical Imaging, Fujian Medical University, Fuzhou 350122, China; Clinical Research Center for Radiology and Radiotherapy of Fujian Province (Digestive, Hematological and Breast Malignancies), Fuzhou 350001, China.
- The Fourth Clinical College, Henan Medical University, Xinxiang, 453003, China.
- Department of Pathology, Fujian Medical University Union Hospital, Fuzhou 350001, China.
- Department of Radiology, The First Affiliated Hospital of Zhengzhou University, Zhengzhou 450052, China; Henan Key Laboratory of Image Diagnosis and Treatment for Digestive System Tumor, Zhengzhou 450052, China; Henan Key Laboratory of CT Imaging, Zhengzhou 450052, China.
- Department of Gastric Surgery, Fujian Medical University Union Hospital, Fuzhou 350001, China.
- Department of Radiology, Xinxiang Central Hospital, The Fourth Clinical College, Henan Medical University, Xinxiang, 453003, China.
- Department of Radiology, Fujian Medical University Union Hospital, Fuzhou 350001, China.
- Department of Radiology, Fujian Medical University Union Hospital, Fuzhou 350001, China; The School of Medical Imaging, Fujian Medical University, Fuzhou 350122, China; Fujian Key Laboratory of Intelligent Imaging and Precision Radiotherapy for Tumors (Fujian Medical University), Fuzhou 350001, China. Electronic address: [email protected].
- School of Computer and Data Science, Minjiang University, Fuzhou 350108, China. Electronic address: [email protected].
- Department of Radiology, Fujian Medical University Union Hospital, Fuzhou 350001, China; The School of Medical Imaging, Fujian Medical University, Fuzhou 350122, China; Fujian Key Laboratory of Intelligent Imaging and Precision Radiotherapy for Tumors (Fujian Medical University), Fuzhou 350001, China. Electronic address: [email protected].
Abstract
Perineural invasion (PNI) is crucial for risk stratification and treatment planning. This study aims to develop a novel multimodal model for preoperative prediction of PNI in gastric cancer. A total of 250 patients with pathologically confirmed gastric cancer from multiple centers were enrolled and divided into a training cohort (n = 138), an internal validation cohort (n = 59), and an external validation cohort (n = 53). Spectral CT parameters from both intra- and peritumoral regions were acquired using the GSI viewer software, while deep learning features were extracted via a ResNet-50 architecture. Clinical features were screened through univariate and multivariate regression analyses. Based on machine learning, we developed a spectral parameter model, a deep learning feature model, a clinical prediction model, and a multimodal fusion model (MFM). The predictive performance of the models was evaluated using calibration curves, receiver operating characteristic curves, and decision curve analysis (DCA). Model interpretability was achieved by employing Shapley Additive exPlanations (SHAP). Compared with individual models, MFM demonstrated the best predictive performance, with area under the curve (AUC) values of 0.926, 0.885, and 0.871 in the training, internal validation, and external validation cohorts, respectively. DCA confirmed that MFM provided a higher net benefit over a wide range of threshold probability. Calibration curve showed MFM had better predictive consistency, and the Hosmer-Lemeshow test indicated good fit across all cohorts (all p-values > 0.05). SHAP effectively interpreted the decision-making process of the model. The novel multimodal fusion model, which integrates clinical, spectral CT parameters, and deep learning features from intra- and peritumoral regions, demonstrated outstanding performance in preoperatively predicting PNI in gastric cancer.