From the Clinical Research Center (J.W.) and Institute of Diagnostic and Interventional Radiology, Department of Radiology (X.S., X.W., L.D., Z.S., Y. Zhu, Y.L.), Shanghai Sixth People's Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, No. 600 Yi Shan Rd, Shanghai 200233, China; School of Health Science and Engineering, University of Shanghai for Science and Technology, Shanghai, China (J.W., M.W., Y.J.); Shukun (Beijing) Network Technology, Beijing, China (Zhiwen Yang, C.M.); Department of Radiology, The First Affiliated Hospital of Soochow University, Jiangsu, China (C.H.); Department of Radiology, Shanghai General Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China (X.X.); Department of Cardiology, Beijing Friendship Hospital of Capital Medical University, Beijing, China (Zhenghan Yang); Department of Radiology, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, Henan, China (Y. Zhang); Department of Radiology, The First Affiliated Hospital of Chongqing Medical University, Chongqing, China (F.L.); and Department of Nuclear Medicine, Xuanwu Hospital, Capital Medical University, Beijing, China (J.L.).
Radiology. 2024 Aug;312(2):e233197. doi: 10.1148/radiol.233197.
Background Deep learning (DL) could improve the labor-intensive, challenging processes of diagnosing cerebral aneurysms but requires large multicenter data sets. Purpose To construct a DL model using a multicenter data set for accurate cerebral aneurysm segmentation and detection on CT angiography (CTA) images and to compare its performance with radiology reports. Materials and Methods Consecutive head or head and neck CTA images of suspected unruptured cerebral aneurysms were gathered retrospectively from eight hospitals between February 2018 and October 2021 for model development. An external test set with reference standard digital subtraction angiography (DSA) scans was obtained retrospectively from one of the eight hospitals between February 2022 and February 2023. Radiologists (reference standard) assessed aneurysm segmentation, while model performance was evaluated using the Dice similarity coefficient (DSC). The model's aneurysm detection performance was assessed by sensitivity and comparing areas under the receiver operating characteristic curves (AUCs) between the model and radiology reports in the DSA data set with use of the DeLong test. Results Images from 6060 patients (mean age, 56 years ± 12 [SD]; 3375 [55.7%] female) were included for model development (training: 4342; validation: 1086; and internal test set: 632). Another 118 patients (mean age, 59 years ± 14; 79 [66.9%] female) were included in an external test set to evaluate performance based on DSA. The model achieved a DSC of 0.87 for aneurysm segmentation performance in the internal test set. Using DSA, the model achieved 85.7% (108 of 126 aneurysms [95% CI: 78.1, 90.1]) sensitivity in detecting aneurysms on per-vessel analysis, with no evidence of a difference versus radiology reports (AUC, 0.93 [95% CI: 0.90, 0.95] vs 0.91 [95% CI: 0.87, 0.94]; = .67). Model processing time from reconstruction to detection was 1.76 minutes ± 0.32 per scan. Conclusion The proposed DL model could accurately segment and detect cerebral aneurysms at CTA with no evidence of a significant difference in diagnostic performance compared with radiology reports. © RSNA, 2024 See also the editorial by Payabvash in this issue.
背景 深度学习（DL）可以改进诊断脑动脉瘤这一劳动密集且具有挑战性的过程，但需要大型多中心数据集。目的 构建一个使用多中心数据集的 DL 模型，以便在 CT 血管造影（CTA）图像上准确分割和检测脑动脉瘤，并比较其与放射科报告的性能。材料与方法 回顾性收集 2018 年 2 月至 2021 年 10 月期间来自 8 家医院的疑似未破裂脑动脉瘤的连续头部或头颈部 CTA 图像，用于模型开发。使用 2022 年 2 月至 2023 年 2 月期间 8 家医院之一的回顾性获得外部测试集，该测试集使用参考标准数字减影血管造影（DSA）扫描。放射科医生（参考标准）评估了动脉瘤的分割情况，而模型的性能则使用 Dice 相似系数（DSC）进行评估。使用 DeLong 检验比较模型和 DSA 数据集中放射科报告的受试者工作特征曲线（AUC）下面积，评估模型在 DSA 数据集中的动脉瘤检测性能。结果 共纳入 6060 例患者（平均年龄，56 岁±12[标准差]；3375 例[55.7%]为女性）进行模型开发（训练：4342 例；验证：1086 例；内部测试集：632 例）。另有 118 例患者（平均年龄，59 岁±14 岁；79 例[66.9%]为女性）纳入外部测试集，以基于 DSA 评估性能。该模型在内部测试集中的动脉瘤分割性能的 DSC 为 0.87。在基于血管的分析中，该模型在检测动脉瘤方面的敏感性为 85.7%（126 个动脉瘤中的 108 个[95%CI：78.1，90.1]），与放射科报告无差异（AUC，0.93[95%CI：0.90，0.95] vs 0.91[95%CI：0.87，0.94]； =.67）。从重建到检测，模型处理时间为每次扫描 1.76 分钟±0.32 分钟。结论 与放射科报告相比，所提出的 DL 模型在 CTA 上可以准确分割和检测脑动脉瘤，且诊断性能无显著差异。 ©2024 RSNA，见本期 Payabvash 社论。
