Deep learning-based identification and localization of intracranial hemorrhage in patients using a large annotated head computed tomography dataset: A retrospective multicenter study

作者：Jun Liu, Weijie Fan, Yi Yang, Peng Qi, Bingjun Ji, Luping He, Yang Li, Jing Yuan, Wei Li, Xianqi Wang, Yi Wu, Chen Liu, Q X Gong, Mi He, Yeqin Fu, Dong Zhang, Si Zhang, Yongjian Nian · 发表于：Intelligent Medicine · 年份：2024 · DOI：10.1016/j.imed.2024.11.002 · 被引用次数：9 · 研究领域：Intracerebral and Subarachnoid Hemorrhage Research、Acute Ischemic Stroke Management、Machine Learning in Healthcare

Accurately identifying and localizing the five subtypes of intracranial hemorrhage (ICH) are crucial steps for subsequent clinical treatment; however, the lack of a large computed tomography (CT) dataset with annotations of the categorization and localization of ICH considerably limits the development of deep learning-based identification and localization methods. We aimed to construct this large dataset and develop a deep learning-based model to identify and localize the five ICH subtypes, including intraventricular hemorrhage (IVH), intraparenchymal hemorrhage (IPH), subdural hemorrhage (SDH), subarachnoid hemorrhage (SAH), and epidural hemorrhage (EDH), in non-contrast head CT scans. Based on the public Radiological Society of North America (RSNA) 2019 dataset, we constructed a large CT dataset named RSNA 2019+ that was annotated for bleeding localization of the five ICH subtypes by three radiologists. An improved YOLOv8 architecture with the bidirectional feature pyramid network was proposed and trained using the RSNA 2019+ training dataset and evaluated on the RSNA 2019+ test dataset. The public CQ500, and two private datasets collected from the Xinqiao and Sunshine Union Hospitals, respectively, were also annotated to perform multicenter validation. Furthermore, the performance of the deep learning model was compared with that of four radiologists. Multiple performance metrics, including the average precision (AP), precision, recall and F1-score, were used for performan...