文献速递静脉血栓栓塞

人工智能整合静脉血栓栓塞症诊疗:预测模型、实施挑战与未来方向

Integrating artificial intelligence into venous thromboembolism care: Predictive models, implementation challenges, and future directions.
PubMed 文献 · J Vasc Surg Venous Lymphat Disord · 2026-08-31

本文综述了23项人工智能在静脉血栓栓塞症防治中的应用研究,其中20项预测或检测研究采用PROBAST+AI评估,18项(90.0%)存在高偏倚风险;仅7项(35%)进行了有意义的外部或前瞻性验证,5项仍为高偏倚风险;14项(70.0%)未报告校准。判别力从中等到高不等,但受回顾性设计、内部验证、类别不平衡等因素限制。智能体AI仍处于概念阶段,需人工监督。

要点

  • 现有静脉血栓栓塞症AI预测模型普遍存在高偏倚风险,外部验证不足。
  • 多数研究未报告校准,判别力变异大,过拟合常见。
  • 智能体AI尚处概念阶段,临床应用前需前瞻性多中心评估与标准化偏倚审计。

核心结论

  • 人工智能在静脉血栓栓塞症全程管理中具有潜力,但当前证据受高偏倚风险、验证不足和校准缺失限制,常规临床应用前需严格评估。

研究设计与样本

  • 综述纳入23项研究,其中20项预测或检测研究采用PROBAST+AI评估,3项评估临床影响干预。

主要发现

  • 20项预测或检测研究中18项(90.0%)为高偏倚风险。
  • 仅7项研究(35%)显示有意义的外部或前瞻性验证,其中5项仍为高偏倚风险。
  • 14项研究(70.0%)未报告校准。
  • 判别力评分从中等到高不等,但性能因队列而异,常受回顾性设计、内部验证、类别不平衡、低事件数和过拟合限制。

临床怎么用

  • 当前人工智能模型不宜直接用于常规静脉血栓栓塞症诊疗决策,需谨慎解读并优先选择经过外部验证和校准评估的工具。

局限性

  • 纳入研究多为回顾性设计、内部验证,存在类别不平衡、低事件数和过拟合问题,且多数未报告校准。

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

Artificial intelligence (AI) has the potential to support personalized, multidisciplinary, data-driven care for venous thromboembolism prevention, detection, imaging, management, and follow-up. We reviewed 23 studies; 20 prediction or detection studies were assessed with PROBAST+AI, while 3 evaluated clinical-impact interventions. Of these, 18 of 20 studies (90.0%) were at high risk of bias. Only 7 studies (35%) showed meaningful external or prospective validation, although 5of 7 remained at high risk of bias. Calibration was not reported in 14 studies (70.0%). Discrimination scores ranged from modest to high, but performance varied across cohorts and was often limited by the retrospective design, internal validation, class imbalance, low event counts, and overfitting. Currently, agentic AI remains conceptual and needs human oversight. Future research should emphasize prospective multicenter evaluation, calibration, transparent reporting, regulatory adherence, and standardized bias auditing before AI is used routinely in clinical practice.

来源:J Vasc Surg Venous Lymphat Disord,查看原文。版权归原作者所有。

引用

  1. Moreno O, Zhang R, Huang A, et al. Integrating artificial intelligence into venous thromboembolism care: Predictive models, implementation challenges, and future directions. J Vasc Surg Venous Lymphat Disord. 2026. PMID: 42674218. doi:10.1016/j.jvsv.2026.102608

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。