Preoperative prediction of metachronous liver metastasis in colorectal cancer using a deep learning-based radiomics model with automatic segmentation: a multicenter study - PubMed
5 hours ago
- #Colorectal cancer
- #Deep learning
- #Radiomics
- Developed a deep learning-based radiomics model for predicting metachronous liver metastasis (MLM) in colorectal cancer (CRC).
- Used nnU-Net for automatic segmentation of CT images with high accuracy (CRC DSC: 86.2% validation, 81.0% test; liver DSC: 96.5% validation, 94.7% test).
- Identified AFP, dissected lymph nodes, perineural invasion, and tumor nodules as independent predictors of MLM.
- Combined clinical-radiomic model outperformed individual models (AUC: 0.972/0.875/0.814; C-index: 0.819/0.728/0.690).
- SHAP analysis highlighted lymph node count, AFP, tumor nodules, and multiregional radiomics features as key contributors.
- High-risk patients showed significantly reduced MLM-free survival (log-rank P < 0.0001).
- Model supports personalized treatment planning and optimization of therapeutic strategies.