动脉内治疗（IAT）是治疗不可切除的肝细胞癌（HCC）的有希望的选择。在进行 IAT 之前对预后风险进行分层对于临床决策和设计未来的临床试验非常重要。开发和验证基于机器学习 (ML) 的决策支持模型 (MLDSM)，用于推荐不可切除 HCC 的 IAT 模式。2014 年 10 月之间截至2022年10月，共回顾性纳入13家三级医院2959例接受初次IAT的HCC患者。这些患者被分为训练队列 (n = 1700)、验证队列 (n = 428) 和测试队列 (n = 200)。输入了 32 个临床变量和 5 种监督 ML 算法，包括 eXtreme Gradient Boosting（ XGBoost）、分类梯度提升（CatBoost）、梯度提升决策树（GBDT）、轻梯度提升机（LGBM）和随机森林（RF），使用接受者操作特征曲线（AUC）下的面积与 DeLong 测试进行比较总共1856名患者被分配到单独IAT组（I-A），1103名患者被分配到IAT联合组（I-C）。 I-A 组的 12 个月死亡率为 31.9%（352/1103），I-C 组为 50.4%（936/1856）。对于测试队列，在 I-C 组中，CatBoost 模型在输入 30 个变量时实现了最佳区分，AUC 为 0.776（95% 置信区间 [CI]，0.833-0.868）。在 I-A 组中，LGBM 模型在输入 24 个变量时实现了最佳区分，AUC 为 0.776（95% CI，0.833-0.868）。根据决策树，BCLC分级、局部治疗和直径作为前三个变量用于指导IAT模式之间的临床决策。MLDSM可以准确地对接受IAT的HCC患者的预后风险进行分层，从而帮助医生做出IAT决策并为临床实践中的监测策略提供指导。© 2024。作者，获得 Springer Nature Limited 的独家许可。

Intra-arterial therapies(IATs) are promising options for unresectable hepatocellular carcinoma(HCC). Stratifying the prognostic risk before administering IAT is important for clinical decision-making and for designing future clinical trials.To develop and validate a machine learning(ML)-based decision support model(MLDSM) for recommending IAT modalities for unresectable HCC.Between October 2014 and October 2022, a total of 2,959 patients with HCC who underwent initial IATs were enroled retrospectively from 13 tertiary hospitals. These patients were divided into the training cohort (n = 1700), validation cohort (n = 428), and test cohort (n = 200).Thirty-two clinical variables were input, and five supervised ML algorithms, including eXtreme Gradient Boosting (XGBoost), Categorical Gradient Boosting (CatBoost), Gradient Boosting Decision Tree (GBDT), Light Gradient Boosting Machine (LGBM) and Random Forest (RF), were compared using the areas under the receiver operating characteristic curve (AUC) with the DeLong test.A total of 1856 patients were assigned to the IAT alone Group(I-A), and 1103 patients were assigned to the IAT combination Group(I-C). The 12-month death rates were 31.9% (352/1103) in the I-A group and 50.4% (936/1856) in the I-C group. For the test cohort, in the I-C group, the CatBoost model achieved the best discrimination when 30 variables were input, with an AUC of 0.776 (95% confidence intervals [CI], 0.833-0.868). In the I-A group, the LGBM model achieved the best discrimination when 24 variables were input, with an AUC of 0.776 (95% CI, 0.833-0.868). According to the decision trees, BCLC grade, local therapy, and diameter as top three variables were used to guide clinical decisions between IAT modalities.The MLDSM can accurately stratify prognostic risk for HCC patients who received IATs, thus helping physicians to make decisions about IAT and providing guidance for surveillance strategies in clinical practice.© 2024. The Author(s), under exclusive licence to Springer Nature Limited.