深度学习自动分割评估TEVAR术后随访CT
本研究构建人工智能流程,自动分析胸主动脉腔内修复(TEVAR)术后随访CT血管造影,测量近端与远端锚定区平均直径、支架长度、最大动脉瘤直径及瘤体体积。共纳入45例TEVAR患者,排除3份(6.6%)分割失败的CT,分析84份CT(术前42份、随访42份);结果显示与人工测量一致性良好,且随访中最大动脉瘤直径与瘤体体积显著下降。
要点
- AI流程可自动量化TEVAR锚定区直径与支架长度,与人工测量一致性较好。
- 术后1年随访中最大动脉瘤直径与瘤体体积较术前显著下降。
- 分割失败率需关注,3份CT(6.6%)因分割失败被排除。
核心结论
- 基于AI的流程可可靠、可重复地量化TEVAR术后支架锚定区,与人工测量高度一致,并检测到随访期形态学显著变化。
研究设计与样本
- 开发深度学习网络自动测量TEVAR术后随访CTA的近端锚定区直径、远端锚定区直径、支架长度、最大动脉瘤直径与瘤体体积。
- 纳入45例TEVAR患者;3份CT(6.6%)因分割失败被排除,最终分析84份CT(术前42份、随访42份)。
- 以Dice相似系数评估分割精度,以组内相关系数(ICC)评估与人工测量的一致性;人工测量由经验丰富的血管外科医生使用EndoSize软件基于标准化中心线标志完成。
主要发现
- 术后1个月近端锚定区直径为32 ± 4.7 mm(ICC 0.86),术后1年为33.41 ± 5.8 mm(ICC 0.78)。
- 术后1个月远端锚定区直径为30.15 ± 4.54 mm(ICC 0.97),术后1年为31.35 ± 4.85 mm(ICC 0.83)。
- 支架长度保持稳定(约212 mm,ICC >0.98)。
- 最大动脉瘤直径由57.4 ± 14.8 mm降至55.5 ± 13.2 mm(P < 0.0001)。
- 瘤体体积由68.6 ± 86.1 mm3降至54.3 ± 103.7 mm3(P < 0.0001),最大动脉瘤直径变化与体积变化强相关(r = 0.77,P < 0.0001)。
临床怎么用
- 该AI流程可用于TEVAR术后随访CTA的标准化、客观量化评估,辅助测量锚定区直径、支架长度、最大动脉瘤直径与瘤体体积。
- 临床解读结果时需注意分割失败可能导致部分CT被排除,应结合人工复核。
局限性
- 3份CT(6.6%)因分割失败被排除,可能影响流程的普适性。
- 研究样本为45例TEVAR患者、84份CT,样本量有限。
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
BACKGROUND: The objective of this study was to apply an artificial intelligence (AI) pipeline for the automatic analysis of follow-up computed tomography angiography after thoracic endovascular aortic repair (TEVAR). METHODS: A deep learning network was developed to automatically measure the mean diameters of the proximal (D_LZP) and distal (D_LZD) landing zones, stent length (L), maximum aneurysm diameter (D_MAX), and aneurysm volume. Segmentation accuracy was assessed with the Dice Similarity Coefficient and agreement with manual measurements using the intraclass correlation coefficient (ICC). Manual measurements were obtained by an experienced vascular surgeon using dedicated software (EndoSize), based on standardized centerline landmarks. RESULTS: The study included 45 TEVAR patients; three computed tomography (CT) scans (6.6%) were excluded due to segmentation failure, leaving 84 CT scans for analysis (42 preoperative and 42 follow-up). At 1 month and 1 year, D_LZP was 32 ± 4.7 mm (ICC 0.86) and 33.41 ± 5.8 mm (ICC 0.78), D_LZD 30.15 ± 4.54 mm (ICC 0.97) and 31.35 ± 4.85 mm (ICC 0.83), while stent length remained stable (∼212 mm, ICC >0.98). D_MAX decreased from 57.4 ± 14.8 mm to 55.5 ± 13.2 mm (P < 0.0001), and aneurysm volume decreased from 68.6 ± 86.1 mm3 to 54.3 ± 103.7 mm3 (P < 0.0001), with a strong correlation between changes in D_MAX and volume (r = 0.77, P < 0.0001). CONCLUSION: The proposed AI-based pipeline enables reliable and reproducible quantification of stent-graft landing zones after TEVAR. The strong agreement with manual measurements and the detection of significant morphological changes over time support its potential for standardized, objective, and clinically relevant follow-up.
引用
- Spinella G, Magliocco M, Basso C, et al. Fully Automated Assessment of Post-TEVAR Follow-Up CT Scan Using Deep Learning-Based Segmentation. Ann Vasc Surg. 2026. PMID: 42710832. doi:10.1016/j.avsg.2026.08.031
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。