A radiomics model based on preoperative gadoxetic acid–enhanced magnetic resonance imaging for predicting post-hepatectomy liver failure in patients with hepatocellular carcinoma

Background Post-hepatectomy liver failure (PHLF) is a fatal complication after resection in patients with hepatocellular carcinoma (HCC). It of clinical importance to estimate the risk PHLF preoperatively. Aims This study aimed develop and validate prediction model based on preoperative gadoxetic acid–enhanced magnetic resonance imaging HCC. Methods A total 276 were retrospectively included randomly divided into training test cohorts (194:82). Clinicopathological variables assessed identify significant indicators for prediction. Radiomics features extracted from normal parenchyma at hepatobiliary phase reproducible, robust non-redundant ones filtered modeling. Prediction models developed using clinicopathological (Clin-model), radiomics (Rad-model), their combination. Results The incidence rate was 24% whole cohort. combined model, consisting albumin–bilirubin (ALBI) score, indocyanine green retention 15 min (ICG-R15), Rad-score (derived 16 features) outperformed Clin-model Rad-model. yielded an area under receiver operating characteristic curve (AUC) 0.84 (95% confidence interval (CI): 0.77–0.90) cohort 0.82 CI: 0.72–0.91) demonstrated good consistency by Hosmer–Lemeshow calibration curve. visualized as nomogram estimating individual PHLF. Conclusion combining factors signature can be applied high serve decision aid when planning surgery treatment