Interpretable prediction of macrotrabecular-massive HCC and recurrence-free survival by integrating MRI LI-RADS features with deep learning habitat radiomics.

OBJECTIVE: Macrotrabecular-massive (MTM+) hepatocellular carcinoma (HCC) is associated with poor prognosis and early recurrence. We developed and validated a nomogram integrating magnetic resonance imaging LI-RADS features with deep learning (DL) habitat radiomics for preoperative prediction of MTM + HCC and stratifying patients according to recurrence-free survival (RFS). MATERIALS AND METHODS: In this retrospective multicenter study, 607 patients with early-stage HCC who underwent curative-intent surgical resection (mean age ± standard deviation, 59.6 ± 10.6 years; 474 males, 133 females) were divided into the training (n = 304), internal validation (n = 131), and external (n = 172) test sets. Liver Imaging Reporting and Data System (LI-RADS) features, LI-RADS categorization, and clinical features were analyzed. Finally, a nomogram integrating LI-RADS features, habitat radiomics, and DL features was developed using multivariate logistic regression. Multivariable Cox regression analyses were performed to identify independent prognostic factors. RESULTS: At multivariable analysis, habitat radiomics score, DL score, alpha-fetoprotein (AFP), alanine transferase (ALT), and fat mass were independent predictors of MTM + HCC. The DL habitat radiomics nomogram demonstrated powerful performance with areas under the receiver operating characteristic curve (AUCs) of 0.897 (95% confidence interval [CI]: 0.854-0.940), 0.757 (95% CI: 0.674-0.827), and 0.858 (95% CI: 0.797-0.906) in the training, internal validation, and external test sets, respectively. After multivariable analysis, AFP (p < 0.001) and nomogram-predicted MTM state (p = 0.039) resulted as independent prognostic factors. CONCLUSION: The nomogram integrating LI-RADS features, habitat radiomics, and DL features accurately identified patients with MTM + HCC and stratified patients by RFS. KEY POINTS: Question Preoperative integrative methods for accurate prediction of MTM + HCC and postoperative recurrence risk stratification remain insufficient. Findings The nomogram integrating LI-RADS features and DL habitat radiomics effectively diagnosed MTM + HCC and showed superior diagnostic performance compared to the Clinical-radiologic model. Relevance statement The nomogram provides a robust, non-invasive tool for preoperative identification of MTM + HCC and recurrence risk stratification, facilitating surgical planning and postoperative surveillance strategies.