Preoperative identification of deep myometrial invasion in endometrial cancer: a multicenter MRI study with a vision foundation model-enhanced multimodal deep learning framework.
Authors
Affiliations (13)
Affiliations (13)
- Department of Medical Imaging Center, The First Affiliated Hospital of Jinan University, Guangzhou, China; Department of Radiology, Jiangmen Central Hospital, Jiangmen, China.
- Jiangmen Key Laboratory of Artificial Intelligence in Medical Image Computation and Application, Jiangmen Central Hospital, Jiangmen, China.
- Department of Gynecology, Jiangmen Central Hospital, Jiangmen, China.
- Department of Radiology, Jiangmen Central Hospital, Jiangmen, China.
- Guangxi Key Laboratory of Low-Altitude Unmanned Autonomous Systems, Guilin University of Aerospace Technology, Guilin, China; Laboratory of Intelligent Detection and Information Processing, Guilin University of Aerospace Technology, Guilin, China.
- Laboratory of Intelligent Detection and Information Processing, Guilin University of Aerospace Technology, Guilin, China.
- Department of Radiology, Yuebei People's Hospital, Shaoguan, China.
- Affiliated Dongguan Hospital of Southern Medical University, Dongguan, China.
- Department of Medical Imaging Center, The First Affiliated Hospital of Jinan University, Guangzhou, China.
- The Third Affiliated Hospital of Guangzhou Medical University, Guangzhou, China.
- Department of Radiology, Maoming People's Hospital, Maoming, China.
- Kaiping Central Hospital, Kaiping, China.
- Department of Medical Imaging Center, The First Affiliated Hospital of Jinan University, Guangzhou, China; Jiangmen Key Laboratory of Artificial Intelligence in Medical Image Computation and Application, Jiangmen Central Hospital, Jiangmen, China. Electronic address: [email protected].
Abstract
To develop and validate a Vision Foundation Model-enhanced Multimodal Deep Learning Radiomics (VFM-MDLR) framework that integrates MR imaging with clinicopathological information for noninvasive prediction of deep myometrial invasion (DMI) in patients with endometrial cancer (EC). This retrospective, multicenter study was conducted across seven independent centers and included 1376 EC patients. We developed a VFM-MDLR model based on two MRI sequences for the prediction of DMI. The framework was designed to first employ a General Knowledge Transfer across Heterogeneous-model (GKTH) subnetwork, which adaptively extracts general representations from vision foundation model (VFM). Building on this foundation, a Cross-Sequence Guided Attention (CSGA) module was incorporated to exploit the complementary information between CE-T1WI and T2WI, thereby achieving semantic alignment and synergistic feature representation. These features were then used to derive a deep-learning signature, termed VFM-enhanced Dual-sequence Knowledge Fusion (VFM-DKF), which was further integrated with key clinicopathological variables to construct the final VFM-MDLR predictor. Model performance was systematically evaluated in the internal validation cohort and four external cohorts, while interpretability was assessed using Grad-CAM and SHAP analyses. The VFM-MDLR model, which comprised age, histopathologic grade, the maximum tumor diameter (TMD), and the DLS, demonstrated the best predictive performance, with the highest AUC (0.832-0.877) across all cohorts. No significant difference was observed in performance for DMI detection between the VFM-MDLR model and experienced radiologists' readings (P = 0.915). The proposed VFM-MDLR model showed favorable accuracy in identifying DMI, potentially providing clinicians with a tool to facilitate individualized surgical treatment for patients with EC.