Background: Currently, surgical decisions for hepatocellular carcinoma (HCC) resection are difficult and not sufficiently personalized. We aimed to develop and validate data driven prediction models to assist surgeons in selecting the optimal surgical procedure for patients. Methods: Retrospective data from 361 HCC patients who underwent radical resection in two institutions were included. End-to-end deep learning models were built to automatically segment lesions from the arterial phase (AP) of preoperative dynamic contrast enhanced magnetic resonance imaging (DCE-MRI). Clinical baseline characteristics and radiomic features were rigorously screened. The effectiveness of radiomic features and radiomic-clinical features was also compared. Three ensemble learning models were proposed to perform the surgical procedure decision and the overall survival (OS) and recurrence-free survival (RFS) predictions after taking different solutions, respectively. Results: SegFormer performed best in terms of automatic segmentation, achieving a Mean Intersection over Union (mIoU) of 0.8860. The five-fold cross-validation results showed that inputting radiomic-clinical features outperformed using only radiomic features. The proposed models all outperformed the other mainstream ensemble models. On the external test set, the area under the receiver operating characteristic curve (AUC) of the proposed decision model was 0.7731, and the performance of the prognostic prediction models was also relatively excellent. The application web server based on automatic lesion segmentation was deployed and is available online. Conclusions: In this study, we developed and externally validated the surgical decision-making procedures and prognostic prediction models for HCC for the first time, and the results demonstrated relatively accurate predictions and strong generalizations, which are expected to help clinicians optimize surgical procedures.

Objectives: Intrahepatic cholangiocarcinoma (iCCA) is a highly malignant cancer. More than 70% of patients are diagnosed at an advanced stage. The aim of this study was to evaluate the diagnostic value of plasma miR-21, miR-122, and CA19-9, hoping to establish a novel model to improve the accuracy for diagnosing iCCA. Materials and methods: Plasma miR-21 and miR-122 were detected in 359 iCCA patients and 642 controls (healthy, benign liver lesions, other malignant liver tumors). All 1001 samples were allocated to training cohort (n = 668) and validation cohort (n = 333) in a chronological order. A logistic regression model was applied to combine these markers. Area under the receiver operating characteristic curve (AUC) was used as an accuracy index to evaluate the diagnostic performance. Results: Plasma miR-21 and miR-122 were significantly higher in iCCA patients than those in controls. Higher plasma miR-21 level was significantly correlated with larger tumor size (p = 0.030). A three-marker model was constructed by using miR-21, miR-122 and CA19-9, which showed an AUC of 0.853 (95% CI: 0.824–0.879; sensitivity: 73.0%, specificity: 87.4%) to differentiate iCCA from controls. These results were subsequently confirmed in the validation cohort with an AUC of 0.866 (0.825–0.901). The results were similar for diagnosing early (stages 0–I) iCCA patients (AUC: 0.848) and CA19-9^negative iCCA patients (AUC: 0.795). Conclusions: We established a novel three-marker model with a high accuracy based on a large number of participants to differentiate iCCA from controls. This model showed a great clinical value especially for the diagnosis of early iCCA and CA19-9^negative iCCA.