From the Biomedical Imaging Analysis Group, Computer Science (L.C., D.R.), and Division of Brain Sciences (L.C., A.L.C.J., A.G., J.G., N.M., A.D., B.A., A.M., P.B.), Imperial College London, Charing Cross Hospital, Fulham Palace Rd, 10L21, London W6 8RF, England; Centre for Clinical Brain Sciences, University of Edinburgh, Edinburgh, Scotland (G.M., J.W.); and Department of Radiology, Northwick Park Hospital, London North West Healthcare NHS Trust, London, England (R.P., D.C.).
Radiology. 2018 Aug;288(2):573-581. doi: 10.1148/radiol.2018171567. Epub 2018 May 15.
Purpose To validate a random forest method for segmenting cerebral white matter lesions (WMLs) on computed tomographic (CT) images in a multicenter cohort of patients with acute ischemic stroke, by comparison with fluid-attenuated recovery (FLAIR) magnetic resonance (MR) images and expert consensus. Materials and Methods A retrospective sample of 1082 acute ischemic stroke cases was obtained that was composed of unselected patients who were treated with thrombolysis or who were undergoing contemporaneous MR imaging and CT, and a subset of International Stroke Thrombolysis-3 trial participants. Automated delineations of WML on images were validated relative to experts' manual tracings on CT images, and co-registered FLAIR MR imaging, and ratings were performed by using two conventional ordinal scales. Analyses included correlations between CT and MR imaging volumes, and agreements between automated and expert ratings. Results Automated WML volumes correlated strongly with expert-delineated WML volumes at MR imaging and CT (r = 0.85 and 0.71 respectively; P < .001). Spatial-similarity of automated maps, relative to WML MR imaging, was not significantly different to that of expert WML tracings on CT images. Individual expert WML volumes at CT correlated well with each other (r = 0.85), but varied widely (range, 91% of mean estimate; median estimate, 11 mL; range of estimated ranges, 0.2-68 mL). Agreements (κ) between automated ratings and consensus ratings were 0.60 (Wahlund system) and 0.64 (van Swieten system) compared with agreements between individual pairs of experts of 0.51 and 0.67, respectively, for the two rating systems (P < .01 for Wahlund system comparison of agreements). Accuracy was unaffected by established infarction, acute ischemic changes, or atrophy (P > .05). Automated preprocessing failure rate was 4%; rating errors occurred in a further 4%. Total automated processing time averaged 109 seconds (range, 79-140 seconds). Conclusion An automated method for quantifying CT cerebral white matter lesions achieves a similar accuracy to experts in unselected and multicenter cohorts.
目的 通过与液体衰减恢复（FLAIR）磁共振成像（MR）和专家共识比较，验证一种在急性缺血性卒中多中心队列中对 CT 图像上脑白质病变（WML）进行分割的随机森林方法。
材料与方法 本回顾性研究纳入了 1082 例急性缺血性卒中患者，包括接受溶栓治疗的患者和同时接受 MR 成像和 CT 检查的患者，以及国际卒中溶栓 3 期试验的部分参与者。将 WML 的自动勾画与 CT 图像上的专家手动勾画进行比较，采用两种常规等级量表进行评估。分析包括 CT 和 MR 成像体积之间的相关性，以及自动勾画和专家评估之间的一致性。
结果 自动勾画的 WML 体积与 MR 成像和 CT 上专家勾画的 WML 体积具有很强的相关性（r 值分别为 0.85 和 0.71；P<0.001）。自动勾画图谱与 WML MR 成像之间的空间相似性与 CT 图像上专家勾画的 WML 轨迹之间没有显著差异。CT 上单个专家的 WML 体积彼此之间相关性良好（r = 0.85），但差异很大（范围为平均值的 91%；中位数估计值为 11 mL；估计范围范围为 0.2-68 mL）。自动评分与共识评分之间的一致性（κ）分别为 0.60（Wahlund 系统）和 0.64（van Swieten 系统），而两个评分系统中专家两两之间的一致性分别为 0.51 和 0.67（P<0.01 Wahlund 系统比较）。准确性不受已建立的梗死、急性缺血性改变或萎缩的影响（P>0.05）。自动预处理失败率为 4%；另外 4%出现评分错误。平均自动处理时间为 109 秒（范围，79-140 秒）。
结论 在未选择和多中心队列中，一种用于量化 CT 脑白质病变的自动方法与专家具有相似的准确性。
