Huan Xinyue, Yang Yang, Niu Shengwen, Yang Yongwei, Tian Bitong, Guo Dajing, Li Kunhua
Department of Radiology, The Second Affiliated Hospital of Chongqing Medical University & Chongqing Medical Imaging Artificial Intelligence Laboratory, No. 74 Linjiang Rd, Yuzhong District, 400010, Chongqing, China.
Department of Radiology, Banan Hospital of Chongqing Medical University, No. 659 Yunan Avenue, Longzhouwan Street, Banan District, 401320, Chongqing, China.
Clin Neuroradiol. 2024 Dec 3. doi: 10.1007/s00062-024-01464-6.
To evaluate the performance of an artificial intelligence (AI) algorithm for automated quantification of arterial stenosis in head and neck CT angiography (CTA).
Patients who received head and neck CTA and DSA between January 2019 and December 2021 in two centers were included. The quantitative performance of CerebralDoc per-lesion was evaluated through intraclass correlation coefficients (ICCs) and Bland-Altman analysis, comparing automated stenosis measurements and manual measurements across 0-100%, < 50%, ≥ 50% and ≥ 70% thresholds. Sensitivity analysis included linear and logistic regression, and subgroups analysis was performed to identify influencing factors.
287 patients with 1765 lesions were analyzed. ICCs between CerebralDoc and DSA for ≥ 50% and ≥ 70% stenosis were excellent (0.955, 0.922, respectively), for 0-100% stenosis was good (0.735), and for < 50% stenosis was poor (0.056). For ≥ 50% and ≥ 70% stenosis of CerebralDoc and CTA manual measurements versus DSA, ICCs were close (0.955 vs 0.994; 0.922 vs 0.986), and differences were small (0.258% vs -0.362%; 0.369% vs -0.199%). The sensitivity analysis revealed that specific locations (V1, V2, V3, V4) and slender vessels have large or remarkable differences ranging from 15.551% to 44.238%.
CerebralDoc exhibited excellent performance in automatically quantifying arterial stenosis of ≥ 50% and ≥ 70% in head and neck CTA. However, further research was needed to improve its performance for < 50% stenosis and to address differences in specific locations and slender vessels.
评估一种人工智能（AI）算法在头颈部CT血管造影（CTA）中自动定量动脉狭窄的性能。
纳入2019年1月至2021年12月期间在两个中心接受头颈部CTA和DSA检查的患者。通过组内相关系数（ICC）和Bland-Altman分析评估CerebralDoc每个病变的定量性能，比较自动狭窄测量值与手动测量值在0-100%、<50%、≥50%和≥70%阈值范围内的差异。敏感性分析包括线性和逻辑回归，并进行亚组分析以确定影响因素。
分析了287例患者的1765个病变。CerebralDoc与DSA对于≥50%和≥70%狭窄的ICC分别为优秀（分别为0.955、0.922），对于0-100%狭窄为良好（0.735），对于<50%狭窄为较差（0.056）。对于CerebralDoc和CTA手动测量值与DSA的≥50%和≥70%狭窄，ICC接近（0.955对0.994；0.922对0.986），差异较小（0.258%对-0.362%；0.369%对-0.199%）。敏感性分析显示，特定部位（V1、V2、V3、V4）和纤细血管的差异较大或显著，范围为15.551%至44.238%。
CerebralDoc在自动定量头颈部CTA中≥50%和≥70%的动脉狭窄方面表现出色。然而，需要进一步研究以提高其对<50%狭窄的性能，并解决特定部位和纤细血管的差异问题。
