文献速递肝胆胰外科围术期管理与 ERAS

人工智能自动生成肝脏虚拟模型并预测肝切除复杂度

Automatic Generation of Liver Virtual Models With Artificial Intelligence: Application to Liver Resection Complexity Prediction.
PubMed 文献 · Ann Surg · 2025-04-11

本研究利用深度学习从术前CT自动生成肝脏三维模型,并构建肝中央区解剖参考框架,训练AI模型预测术中肝切除复杂度(LRC)。在145例HCC患者队列中,AI预测的准确度和AUC分别为79.4±3.4%和85.1±3.2%,优于外科医生个体及综合预判。

要点

  • AI自动生成的肝中央区模型在多种非典型血管解剖中dice评分达82±4.6%。
  • AI预测术中LRC的准确度79.4±3.4%、AUC 85.1±3.2%,优于外科医生预判。
  • 该工具可依据预测LRC水平帮助患者导向合适的医疗中心。

核心结论

  • 自动数字化工具能准确预测术中肝切除复杂度,为肿瘤外科手术规划提供新途径。

研究设计与样本

  • 基于深度学习从患者CT扫描生成器官、肿瘤和血管三维模型,并训练AI模型预测术中LRC。
  • 在145例HCC患者队列中运行自动流程,并将AI预测与仅查看术前CT的外科医生预判进行比较。

主要发现

  • 三维重建在基准数据集上成功评估。
  • 肝中央区在多种非典型血管解剖中准确生成,dice评分82±4.6%。
  • 自动流程在145例HCC患者队列中成功运行。
  • AI预测LRC的准确度和AUC分别为79.4±3.4%和85.1±3.2%,优于外科医生个体及综合预判。

临床怎么用

  • 可利用AI预测的LRC水平将患者导向合适的医疗中心。
  • 该工具可辅助肿瘤外科手术规划。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

OBJECTIVE: The clinical aim of this work is to predict intraoperative LRC from preoperative CT scans only. BACKGROUND: Liver resection (LR) is the most prevalent curative treatment for primary liver cancer, yet overall mortality/morbidity rates remain elevated. The conventional definition and classification of LR complexity (LRC) lack the inclusion of the disease-induced 3D anatomic surgery complexity. METHODS: 3D models of the organ, tumors, and blood vessels were generated from deep learning models trained on patients' CT scans. The surgeons' expertise on which anatomic factors lead to LRC was translated into a new anatomic frame of reference around the Hepatic Central Zone (HCZ). A fully automatic pipeline to generate the HCZ and quantify the tumor position relative to it was assessed. An AI model was then trained to predict LRC from a patient cohort for whom LRC was annotated at the end of each surgery. The AI prediction was finally compared to the prediction of surgeons who only saw the patient's preoperative CT scan. RESULTS: The 3D reconstructions are successfully evaluated on benchmark data sets. The HCZ is accurately generated for a variety of atypical vascular anatomies (dice score 82±4.6%). The automatic pipeline is successfully run on a 145 HCC patient cohort. The predicted LRC outperforms the surgeons' individual and combined anticipated complexities (accuracy and AUC scores: 79.4±3.4% and 85.1±3.2%, respectively). CONCLUSIONS: This automatic digital tool accurately predicts intraoperative LRC and paves the way for an innovative oncology surgery planning. This tool could help orient patients toward appropriate medical centers depending on the predicted LRC level.

来源:Ann Surg,查看原文。版权归原作者所有。

引用

  1. Ali O, Bône A, Accardo C, et al. Automatic Generation of Liver Virtual Models With Artificial Intelligence: Application to Liver Resection Complexity Prediction. Ann Surg. 2025. PMID: 40211963. doi:10.1097/SLA.0000000000006722

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。