Radiomics Integrated with Deep Learning Based on Contrast-Enhanced MRI for Predicting Short-Term Efficacy of…

AI / informaticsJournal of imaging informatics in medicine· Aug 2026

Radiomics Integrated with Deep Learning Based on Contrast-Enhanced MRI for Predicting Short-Term Efficacy of Drug-Eluting Beads Transarterial Chemoembolization in Hepatocellular Carcinoma from Multicenter Retrospective Study.

Zihan Xi, Yuchi Tian, Lulu Liu … Yongbo Yang

Interventional RadiologyAI / informaticsQ1IF 5.3

Combined radiomics-deep learning model predicts drug-eluting bead transarterial chemoembolization response in liver cancer

Radiomics Integrated with Deep Learning Based on Contrast-Enhanced MRI for Predicting Short-Term Efficacy of Drug-Eluting Beads Transarterial Chemoembolization in Hepatocellular Carcinoma from Multicenter Retrospective Study.

Zihan Xi … Yongbo Yang

Journal of imaging informatics in medicine · Aug 2026 · Q1 in Radiology

Purpose

We aimed to develop and validate a predictive model combining radiomics, deep learning, and clinical features for the preoperative prediction of the short-term efficacy of initial drug-eluting bead transarterial chemoembolization (drug-eluting bead transarterial chemoembolization) treatment in patients with hepatocellular carcinoma (hepatocellular carcinoma).

Methods

This retrospective cohort study included 113 internal and 32 external patients from three centers with intermediate and advanced hepatocellular carcinoma were included.

n = 113 internal and 32 external

Results

External test accuracy predicting treatment response; AUC 1.0 is perfect, 0.5 is chance.

0.864AUC

Training set

0.822AUC

Internal testing set

0.804AUC

External testing set

More results

We extracted 1496 radiomics features per sequence based on three sequences: arterial phase (arterial phase), diffusion-weighted imaging (diffusion-weighted imaging), and T2-weighted imaging (T2WI) based on contrast-enhanced magnetic resonance imaging.

More results

Specifically, the activation of 512 neurons from the final fully connected layer of the network was employed as a deep learning feature.

A total of 34 radiomics features were selected using least absolute shrinkage and selection operator regression.

“

Conclusion

The comprehensive model based on radiomics, deep learning, and clinical features can effectively predict short-term efficacy of initial drug-eluting bead transarterial chemoembolization in patients with hepatocellular carcinoma.

Radiomics Integrated with Deep Learning Based on Contrast-Enhanced MRI for Predicting Short-Term Efficacy of Drug-Eluting Beads Transarterial Chemoembolization in Hepatocellular Carcinoma from Multicenter Retrospective Study.

Radiomics Integrated with Deep Learning Based on Contrast-Enhanced MRI for Predicting Short-Term Efficacy of Drug-Eluting Beads Transarterial Chemoembolization in Hepatocellular Carcinoma from Multicenter Retrospective Study.

Related papers