JoVE - Journal of Visualized Experiments
JoVE Visualize
Contact Us

Integrating CEUS Imaging Features and LI-RADS Classification for Postoperative Early Recurrence Prediction in Solitary Hepatocellular Carcinoma: A Machine Learning-Based Prognostic Approach

  • 0Department of Medical Ultrasound, The First Affiliated Hospital of Guangxi Medical University, Nanning, Guangxi Zhuang Autonomous Region, People's Republic of China.
Journal of Hepatocellular Carcinoma +

|

July 9, 2025

|

July 9, 2025

View abstract on PubMed

Summary

This summary is machine-generated.

This study developed a machine learning model using contrast-enhanced ultrasound (CEUS) and LI-RADS classification to predict early hepatocellular carcinoma (HCC) recurrence after surgery. The Gradient Boosting Machine (GBM) model showed strong predictive performance, aiding in post-operative management.

Area Of Science

  • Medical Imaging
  • Machine Learning in Oncology
  • Hepatocellular Carcinoma Research

Background

  • Early postoperative recurrence of hepatocellular carcinoma (HCC) significantly impacts patient survival and treatment planning.
  • Accurate prediction of recurrence is crucial for optimizing surveillance strategies and patient management.
  • Integrating imaging features with machine learning offers a promising avenue for improving prognostic accuracy.

Purpose Of The Study

  • To develop and validate a machine learning (ML) model for predicting early postoperative recurrence in HCC patients.
  • To integrate contrast-enhanced ultrasound (CEUS) features with Liver Imaging Reporting and Data System (LI-RADS) classification for enhanced prediction.
  • To assess the performance of various ML algorithms in predicting HCC recurrence.

Main Methods

  • Retrospective analysis of 279 HCC patients who underwent surgical resection.
  • Integration of CEUS-derived features, LI-RADS classification, and clinical-pathological variables.
  • Development and comparison of four ML models (RSF, GBM, CoxBoost, XGBoost) using a 7:3 training-validation split.
  • Performance evaluation using C-index, AUC, calibration curves, DCA, and KM survival analysis.

Main Results

  • Five significant features identified: microvascular invasion (MVI), tumor size, LI-RADS classification, tumor necrosis, and arterial enhancement patterns.
  • The Gradient Boosting Machine (GBM) model demonstrated the best performance, achieving high C-index and AUC values in both training and validation cohorts.
  • LI-RADS classification, MVI, and tumor size were identified as key prognostic indicators.
  • Kaplan-Meier analysis confirmed the model's ability to stratify patients into distinct risk groups.

Conclusions

  • The GBM-based ML model integrating CEUS imaging features and LI-RADS classification shows significant potential for predicting early HCC recurrence.
  • This predictive model can assist clinicians in guiding personalized follow-up strategies for HCC patients post-surgery.
  • The study highlights the value of combining advanced imaging analysis with machine learning for improved oncological outcomes.

Keywords: