深度学习全自动容积分割评估胸主动脉、内脏腹主动脉及内脏血管
FAVS的容积分割指标与PCMS相当,中位DSC为0.93(IQR 0.91, 0.94),并正确识别99.5%(183/184)的内脏血管。FAVS可提供与专家手动分割相当的准确、高效分割。
要点
- FAVS容积分割中位DSC为0.93(IQR 0.91, 0.94),与医师间比较相当。
- FAVS正确识别99.5%(183/184)的内脏血管。
- FAVS与PCMS近端内脏直径平均差异-0.11 ± 5.23 mm,医师间为0.03 ± 5.27 mm。
核心结论
- FAVS对胸主动脉和内脏主动脉提供准确、高效的分割,性能与专家医师手动分割相当。
研究设计与样本
- 比较FAVS与PCMS,并评估PCMS的观察者间和观察者内可靠性。
主要发现
- FAVS容积分割中位DSC为0.93(IQR 0.91, 0.94),JI为0.87(IQR 0.84, 0.89),敏感度为0.99(IQR 0.98, 0.99),特异度为1.00(IQR 1.00, 1.00)。
- 医师间比较中位DSC为0.93(IQR 0.86, 0.93),JI为0.87(IQR 0.76, 0.88),敏感度为0.90(IQR 0.86, 0.94),特异度为1.00(IQR 1.00, 1.00)。
- FAVS正确识别99.5%(183/184)的内脏血管。
- FAVS与PCMS分支点坐标差异在CTA空间分辨率范围内:Δx -0.33(IQR -1.70, 1.12)mm,Δy 0.61(IQR -1.25, 3.60)mm,Δz 2.10(IQR 0.37, 5.06)mm。
- 近端内脏直径测量BA-LoA:FAVS vs. PCMS平均差异-0.11 ± 5.23 mm,医师间变异性0.03 ± 5.27 mm。
临床怎么用
- FAVS可用于复杂腹主动脉及胸腹主动脉瘤的监测与治疗规划,提升临床工作流程效率。
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: Computed tomography angiography (CTA) imaging is essential to evaluate and analyse complex abdominal and thoraco-abdominal aortic aneurysms. However, CTA analyses are labour intensive, time consuming, and prone to interphysician variability. Fully automatic volume segmentation (FAVS) using artificial intelligence with deep learning has been validated for infrarenal aorta imaging but requires further testing for thoracic and visceral aorta segmentation. This study assessed FAVS accuracy against physician controlled manual segmentation (PCMS) in the descending thoracic aorta, visceral abdominal aorta, and visceral vasculature. METHODS: This was a retrospective, multicentre, observational cohort study. Fifty pre-operative CTAs of patients with abdominal aortic aneurysm were randomly selected. Comparisons between FAVS and PCMS and assessment of inter- and intra-observer reliability of PCMS were performed. Volumetric segmentation performance was evaluated using sensitivity, specificity, Dice similarity coefficient (DSC), and Jaccard index (JI). Visceral vessel identification was compared by analysing branchpoint coordinates. Bland-Altman limits of agreement (BA-LoA) were calculated for proximal visceral diameters (excluding duplicate renals). RESULTS: FAVS demonstrated performance comparable with PCMS for volumetric segmentation, with a median DSC of 0.93 (interquartile range [IQR] 0.91, 0.94), JI of 0.87 (IQR 0.84, 0.89), sensitivity of 0.99 (IQR 0.98, 0.99), and specificity of 1.00 (IQR 1.00, 1.00). These metrics are similar to interphysician comparisons: median DSC 0.93 (IQR 0.86, 0.93), JI 0.87 (IQR 0.76, 0.88), sensitivity 0.90 (IQR 0.86, 0.94), and specificity 1.00 (IQR 1.00, 1.00). FAVS correctly identified 99.5% (183/184) of visceral vessels. Branchpoint coordinates for FAVS and PCMS were within the limits of CTA spatial resolution (Δx -0.33 [IQR -1.70, 1.12], Δy 0.61 [IQR -1.25, 3.60], Δz 2.10 [IQR 0.37, 5.06] mm). BA-LoA for proximal visceral diameter measurements showed reasonable agreement: FAVS vs. PCMS mean difference -0.11 ± 5.23 mm compared with interphysician variability of 0.03 ± 5.27 mm. CONCLUSION: FAVS provides accurate, efficient segmentation of the thoracic and visceral aorta, delivering performance comparable to manual segmentation by expert physicians. This technology may enhance clinical workflows for monitoring and planning treatments for complex abdominal and thoraco-abdominal aortic aneurysms.
引用
- Pouncey AL, Charles E, Bicknell CD, et al. Editor's Choice - Fully Automatic Volume Segmentation Using Deep Learning Approaches to Assess the Thoracic Aorta, Visceral Abdominal Aorta, and Visceral Vasculature. Eur J Vasc Endovasc Surg. 2025. PMID: 40812505. doi:10.1016/j.ejvs.2025.08.025
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。