Radiomics analysis of R2* maps to predict early recurrence of single hepatocellular carcinoma after hepatectomy

Area of Science:

• Radiology
• Oncology
• Medical Imaging Analysis

Background:

• Hepatocellular carcinoma (HCC) recurrence after partial hepatectomy poses a significant clinical challenge.
• Accurate prediction of early recurrence (ER) is crucial for timely intervention and improved patient outcomes.
• Radiomics analysis, particularly with R2* maps, shows promise in non-invasively assessing tumor characteristics.

Purpose of the Study:

• To evaluate the effectiveness of radiomics analysis using R2* maps for predicting ER in single HCC after partial hepatectomy.
• To develop and validate a predictive model incorporating radiomic features and clinical factors for ER risk stratification.
• To assess the clinical utility of the developed nomogram for personalized treatment decisions in HCC patients.

Main Methods:

• Retrospective analysis of 202 HCC patients undergoing preoperative MRI.
• Feature selection using LASSO regularization to build a radiomic score (Rad-score).
• Development and validation of a predictive nomogram integrating Rad-score, microvascular invasion (MVI), and alpha-fetoprotein (AFP) levels.

Main Results:

• The radiomic score, MVI, and AFP > 400 ng/mL were identified as independent predictors of ER.
• The nomogram achieved an AUC of 0.901 (training) and 0.827 (validation).
• The nomogram demonstrated high predictive performance with an F1 score of 0.831 (training) and 0.808 (validation).

Conclusions:

• A nomogram integrating radiomics, MVI, and AFP effectively predicts early recurrence in HCC.
• This tool facilitates personalized risk classification and informs therapeutic decision-making for HCC patients.
• Radiomics analysis with R2* maps offers a valuable approach for predicting HCC recurrence.