Construction and Validation of a Prediction Model for TACE Resistance in Hepatocellular Carcinoma Based on Contrast-Enhanced Ultrasound Radiomics

Abstract Objective To construct and validate a contrast-enhanced ultrasound (CEUS) radiomics-clinical model for predicting transarterial chemoembolization (TACE) resistance in hepatocellular carcinoma (HCC). Methods A total of 168 HCC patients who underwent CEUS on two vendor platforms (Philips E9 and GE Logiq E95) were split into a training set (n = 118) and an internal validation set (n = 50), with an additional 80 patients forming an external validation cohort. For each patient, 2D regions of interest were manually delineated on the largest cross-sectional slice of each CEUS phase. Radiomic features were extracted and screened by LASSO regression; the twelve retained features were linearly combined into a single Rad-score. The Rad-score and three prespecified clinical risk factors (tumor size, Child-Pugh class, AFP) confirmed by multivariate analysis were combined into a parsimonious logistic regression (LR) model with four predictors in total. Performance was assessed by ROC curve, calibration curve, decision curve analysis (DCA), subgroup analysis, and prospective validation. Results The combined model had AUCs of 0.895, 0.891 and 0.846 in three sets, with subgroup AUCs >0.86. Prospective validation showed TACE plus sorafenib improved 6-month response rate from 26.7% to 53.3% for high-risk patients. Conclusion The CEUS radiomics-clinical model accurately predicts TACE resistance, guiding individualized HCC treatment. Advances in Knowledge This model integrates CEUS multi-phase radiomics and clinical factors, outperforming single clinical/radiomic models. Strict multi-step screening and comprehensive validation avoid data leakage, ensuring robust prediction and clinical translatability.

Authors

Institutions

Publication Details

Journal
British Journal of Radiology
Published
2026-11-14
DOI
https://doi.org/10.1093/bjr/tqag281
Primary Topic
Radiomics and Machine Learning in Medical Imaging
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Construction and Validation of a Prediction Model for TACE Resistance in Hepatocellular Carcinoma Based on Contrast-Enhanced Ultrasound Radiomics

Zhibin Wu, Xiaohui Wang, Lutan Yan, Xiaoli Shi et al.
British Journal of Radiology
Radiomics and Machine Learning in Medical Imaging
article

Construction and Validation of a Prediction Model for TACE Resistance in Hepatocellular Carcinoma Based on Contrast-Enhanced Ultrasound Radiomics

Zhibin Wu, Xiaohui Wang, Lutan Yan, Xiaoli Shi, Lin Zhou
article en

Abstract

Abstract Objective To construct and validate a contrast-enhanced ultrasound (CEUS) radiomics-clinical model for predicting transarterial chemoembolization (TACE) resistance in hepatocellular carcinoma (HCC). Methods A total of 168 HCC patients who underwent CEUS on two vendor platforms (Philips E9 and GE Logiq E95) were split into a training set (n = 118) and an internal validation set (n = 50), with an additional 80 patients forming an external validation cohort. For each patient, 2D regions of interest were manually delineated on the largest cross-sectional slice of each CEUS phase. Radiomic features were extracted and screened by LASSO regression; the twelve retained features were linearly combined into a single Rad-score. The Rad-score and three prespecified clinical risk factors (tumor size, Child-Pugh class, AFP) confirmed by multivariate analysis were combined into a parsimonious logistic regression (LR) model with four predictors in total. Performance was assessed by ROC curve, calibration curve, decision curve analysis (DCA), subgroup analysis, and prospective validation. Results The combined model had AUCs of 0.895, 0.891 and 0.846 in three sets, with subgroup AUCs >0.86. Prospective validation showed TACE plus sorafenib improved 6-month response rate from 26.7% to 53.3% for high-risk patients. Conclusion The CEUS radiomics-clinical model accurately predicts TACE resistance, guiding individualized HCC treatment. Advances in Knowledge This model integrates CEUS multi-phase radiomics and clinical factors, outperforming single clinical/radiomic models. Strict multi-step screening and comprehensive validation avoid data leakage, ensuring robust prediction and clinical translatability.

British Journal of Radiology
First Affiliated Hospital of Zhengzhou University (CN)
Radiomics and Machine Learning in Medical Imaging
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon