What type of study is this?

This is a Quantitative Study study.

October 13, 2025

Individualized Prognostication Based on Deep‐Learning Models Using Computed Tomography as an Imaging Biomarker After Hepatocellular Carcinoma Resection

Key Points

Deep-learning models achieved a predictive accuracy with an area under the curve of 0.70 for recurrence at 2 years and 0.80 for overall survival at 10 years.
Postoperative recurrence rates were 52.6% in high-risk patients versus 18.5% in low-risk patients within 2 years, indicating significant predictive capabilities.
Utilizing computed tomography as an imaging biomarker improved prognostication, aiding in treatment planning for hepatocellular carcinoma patients.
Permutation importance showed that CT imaging analysis had the highest relevance in predicting patient outcomes after surgery.

Abstract

ABSTRACT Introduction No reports described the deep‐learning (DL) models using computed tomography (CT) as an imaging biomarker for predicting postoperative long‐term outcomes in patients with hepatocellular carcinoma (HCC). This study aimed to validate the DL models for individualized prognostication after HCC resection using CT as an imaging biomarker. Methods This study included 1733 patients undergoing hepatic resection for solitary HCC. Participants were classified into training, validation, and test datasets. DL predictive models were developed using clinical variables and CT imaging to predict recurrence within 2 and 5 years and overall survival (OS) of > 5 and > 10 years postoperatively. Youden index was utilized to identify cutoff values. Permutation importance was used to calculate the importance of each explanatory variable. Results DL predictive models for recurrence within 2 and 5 years and OS of > 5 and > 10 years postoperatively were developed in the test datasets, with the area under the curve of 0.70, 0.70, 0.80, and 0.80, respectively. Permutation importance demonstrated that CT imaging analysis revealed the highest importance value. The postoperative recurrence rates within 2 and 5 years were 52.6% versus 18.5% ( p < 0.001) and 78.9% versus 46.7% ( p < 0.001) and overall mortality within 5 and 10 years postoperatively were 45.1% versus 9.2% ( p < 0.001) and 87.1% versus 43.2% ( p < 0.001) in the high‐risk versus low‐risk groups, respectively. Conclusions Our DL models using CT as an imaging biomarker are useful for individualized prognostication and may help optimize treatment planning for patients with HCC.

KI fragen

Bookmark