A retrospective analysis was conducted on preoperative spectral CT images, clinical, and pathological data of 88 patients who underwent radical resection for HCC between 2018 and 2022, categorized into an ER group (n=37) and a non-ER group (n=51). Histogram parameters of the whole tumor were calculated from iodine (water) maps derived from the portal venous phase of spectral CT, including Min, Max, Mean, SD, Variance, Skewness, Kurtosis, Entropy, and percentiles (V10-V95). Multivariate logistic regression analysis was used to assess the predictive factors for NRM-HCC with varying recurrence outcomes, and a comprehensive model was developed. The performance of the comprehensive model was evaluated using receiver operating characteristic (ROC) analysis, calibration curves, and clinical decision curves (DCA).对2018年至2022年期间接受HCC根治性切除术的88例患者的术前光谱CT图像、临床和病理数据进行了回顾性分析，这些患者分为ER组（n=37）和非ER组（n=51）。根据能谱CT门静脉期的碘（水）图计算整个肿瘤的直方图参数，包括最小值、最大值、平均值、SD、方差、偏度、峰度、熵和偏度（V10-V95）。多因素Logistic回归分析用于评估不同复发结局的NRM-HCC的预测因素，并建立了一个综合模型。使用受试者工作特征（ROC）分析、校准曲线和临床决策曲线（DCA）评价综合模型的性能。