Supplementary Material for: Attempt to establish prognostic predictive system for hepatocellular carcinoma using artificial intelligence for assistance with selection of treatment modality

Background/Aim: Because of recent developments in treatments for hepatocellular carcinoma (HCC), methods for determining suitable therapy for initial or recurrent HCC have become important. This study used artificial intelligence (AI) findings to establish a system for predicting prognosis of HCC patients at time of reoccurrence based on clinical data as a reference for selection of treatment modalities.
Material/Methods: As a training cohort, 5701 observations obtained at the initial and each subsequent treatment for recurrence from 1985 HCC patients at a single center from 2000 to 2021 were used. The validation cohort was 5692 observations from patients at multiple centers obtained at the time of the initial treatment. An AI calculating system (PRAID) was constructed based on 25 clinical factors noted at each treatment from the training cohort, then predictive prognostic values for one- and three-year survival in both cohorts were evaluated.
Results: After exclusion of patients lacking clinical data regarding albumin-bilirubin grade (ALBI) and/or tumor-node-metastasis stage (TNM), ALBI-TNM (ALBI-T) and modified ALBI-T scores confirmed that prognosis for patients in both cohorts was similar. The area under the curve (AUC) for prediction of both one- and three-year survival in the validation cohort were 0.841 [sensitivity 0.933 (95%CI 0.925-0.940), specificity 0.517 (95%CI 0.484-0.549)] and 0.796 [sensitivity 0.806 (95%CI 0.790-0.821), specificity 0.646 (95%CI 0.624-0.668)], respectively.
Conclusion: The present PRAID system might provide useful prognostic information related to short and medium survival for decision making regarding best therapeutic modality for both initial and recurrent HCC cases.

背景/目标：鉴于近年来肝细胞癌（HCC）治疗方法的进展，确定初始或复发HCC适宜治疗方案的方法变得至关重要。本研究利用人工智能（AI）的发现，建立了一个基于临床数据预测HCC患者复发时预后的系统，以作为选择治疗方案模式的参考。

材料/方法：作为训练队列，使用了来自2000年至2021年间单个中心1985例HCC患者初次治疗及每次复发治疗的5701个观察值。验证队列包括来自多个中心的5692个初次治疗时的患者观察值。基于训练队列中每次治疗时记录的25个临床因素构建了一个AI计算系统（PRAID），然后评估了两个队列中一岁和三岁生存率的预测预后值。

结果：在排除缺乏关于白蛋白-胆红素分级（ALBI）和/或肿瘤-节点-转移阶段（TNM）临床数据的患者后，ALBI-TNM（ALBI-T）和修改后的ALBI-T评分证实了两个队列中患者的预后相似。验证队列中预测一岁和三岁生存率的曲线下面积（AUC）分别为0.841[敏感性0.933（95%CI 0.925-0.940），特异性0.517（95%CI 0.484-0.549）]和0.796[敏感性0.806（95%CI 0.790-0.821），特异性0.646（95%CI 0.624-0.668）]。

结论：本研究的PRAID系统可能为决策提供有关短期和中长期生存的有用预后信息，以确定初次和复发HCC病例的最佳治疗方案。