Xinpeng Cheng1, Wei Zhang2, Meng Lu Wu3, Nan Jiang1, Zhen Ni Guo1, Xinyi Leng4, Jia Ning Song2, Hang Jin1, Xin Sun1, Fuliang Zhang1, Jing Qin5, Xiuli Yan1, Zhenyu Cai6, Ying Luo6, Yi Yang1, Jia Liu7. 1. Stroke Center, Department of Neurology, Jilin University First Hospital, Changchun, Jilin, CHINA. 2. Institute of Advanced Computing and Digital Engineering, Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences, Shenzhen, Guangdong, CHINA. 3. Institute of Advanced Computing and Digital Engineering, Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences, 1068 Xueyuan Avenue, Xili University Town, Nanshan District, Shenzhen City, Guangdong Province, China, Shenzhen, Guangdong, 518055, CHINA. 4. Department of Medicine and Therapeutics, Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong, HONG KONG. 5. Hong Kong Polytechnic University, Hong Kong, CHINA. 6. Department of Radiology, Fuwai Hospital Chinese Academy of Medical Sciences, Guangdong, CHINA. 7. Shenzhen Institute of Advanced Integration Technology, Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences, Shenzhen, Guangdong, CHINA.
Abstract
BACKGROUND AND PURPOSE: Hematoma expansion is closely associated with adverse functional outcomes in patients with intracerebral hemorrhage (ICH). Prediction of hematoma expansion would therefore be of great clinical significance. We therefore attempted to predict hematoma expansion using a dual-modal machine learning (ML) strategy which combines information from non-contrast computed tomography (NCCT) images and multiple clinical variables. METHOD: We retrospectively identified 140 ICH patients (57 with hematoma expansion) with 5,616 NCCT images of hematoma (2,635 with hematoma expansion) and 10 clinical variables. The dual-modal ML strategy consists of two steps. The first step is to derive a mono-modal predictor from a deep convolutional neural network (DCNN) using solely NCCT images. The second step is to achieve a dual-modal predictor by combining the mono-modal predictor with 10 clinical variables to predict hematoma growth using a multi-layer perception (MLP) network. RESULT: For the mono-modal predictor, the best performance was merely 69.5% in accuracy with solely the NCCT images, whereas the dual-modal predictor could boost the accuracy greatly to be 86.5% by combining clinical variables. CONCLUSION: To our knowledge, this is the best performance from using ML to predict hematoma expansion. It could be potentially useful as a screening tool for high-risk patients with ICH, though further clinical tests would be necessary to show its performance on a larger cohort of patients.
BACKGROUND AND PURPOSE:Hematoma expansion is closely associated with adverse functional outcomes in patients with intracerebral hemorrhage (ICH). Prediction of hematoma expansion would therefore be of great clinical significance. We therefore attempted to predict hematoma expansion using a dual-modal machine learning (ML) strategy which combines information from non-contrast computed tomography (NCCT) images and multiple clinical variables. METHOD: We retrospectively identified 140 ICHpatients (57 with hematoma expansion) with 5,616 NCCT images of hematoma (2,635 with hematoma expansion) and 10 clinical variables. The dual-modal ML strategy consists of two steps. The first step is to derive a mono-modal predictor from a deep convolutional neural network (DCNN) using solely NCCT images. The second step is to achieve a dual-modal predictor by combining the mono-modal predictor with 10 clinical variables to predict hematoma growth using a multi-layer perception (MLP) network. RESULT: For the mono-modal predictor, the best performance was merely 69.5% in accuracy with solely the NCCT images, whereas the dual-modal predictor could boost the accuracy greatly to be 86.5% by combining clinical variables. CONCLUSION: To our knowledge, this is the best performance from using ML to predict hematoma expansion. It could be potentially useful as a screening tool for high-risk patients with ICH, though further clinical tests would be necessary to show its performance on a larger cohort of patients.