Segmentation of Cerebral Small Vessel Diseases-White Matter Hyperintensities Based on a Deep Learning System

被引:4
|
作者
Shan, Wei [1 ,2 ,3 ]
Duan, Yunyun [1 ]
Zheng, Yu [2 ]
Wu, Zhenzhou [2 ]
Chan, Shang Wei [2 ]
Wang, Qun [1 ,2 ,3 ]
Gao, Peiyi [2 ]
Liu, Yaou [2 ]
He, Kunlun [4 ,5 ]
Wang, Yongjun [1 ,2 ]
机构
[1] Capital Med Univ, Beijing Tiantan Hosp, Dept Neurol, Beijing, Peoples R China
[2] Natl Ctr Clin Med Neurol Dis, Beijing, Peoples R China
[3] Beijing Inst Brain Disorders, Beijing, Peoples R China
[4] Chinese Peoples Liberat Army Gen Hosp, Lab Translat Med, Beijing, Peoples R China
[5] Chinese Peoples Liberat Army Gen Hosp, Key Lab Minist Ind & Informat Technol Biomed Engn, Beijing, Peoples R China
基金
中国博士后科学基金;
关键词
masking white matter hyperintensities; deep learning; neural network; segmentation; clinical evaluation; MRI; LESIONS;
D O I
10.3389/fmed.2021.681183
中图分类号
R5 [内科学];
学科分类号
1002 ; 100201 ;
摘要
Objective: Reliable quantification of white matter hyperintensities (WHMs) resulting from cerebral small vessel diseases (CSVD) is essential for understanding their clinical impact. We aim to develop and clinically validate a deep learning system for automatic segmentation of CSVD-WMH from fluid-attenuated inversion recovery (FLAIR) imaging using large multicenter data.Method: A FLAIR imaging dataset of 1,156 patients diagnosed with CSVD associated WMH (median age, 54 years; 653 males) obtained between September 2018 and September 2019 from Beijing Tiantan Hospital was retrospectively analyzed in this study. Locations of CSVD-WMH on the FLAIR scans were manually marked by two experienced neurologists. Using the manually labeled data of 996 patients (development set), a U-shaped novel 2D convolutional neural network (CNN) architecture was trained for automatic segmentation of CSVD-WMH. The segmentation performance of the network was evaluated with per pixel and lesion level dice scores using an independent internal test set (n = 160) and a multi-center external test set (n = 90, three medical centers). The clinical suitability of the segmentation results, classified as acceptable, acceptable with minor revision, acceptable with major revision, and not acceptable, was analyzed by three independent neuroradiologists. The inter-neuroradiologists agreement rate was assessed by the Kendall-W test.Results: On the internal and external test sets, the proposed CNN architecture achieved per pixel and lesion level dice scores of 0.72 (external test set), and they were significantly better than the state-of-the-art deep learning architectures proposed for WMH segmentation. In the clinical evaluation, neuroradiologists observed the segmentation results for 95% of the patients were acceptable or acceptable with a minor revision.Conclusions: A deep learning system can be used for automated, objective, and clinically meaningful segmentation of CSVD-WMH with high accuracy.
引用
收藏
页数:8
相关论文
共 50 条
  • [1] Automatic segmentation of white matter hyperintensities and correlation analysis for cerebral small vessel disease
    Xu, Bin
    Zhang, Xiaofeng
    Tian, Congyu
    Yan, Wei
    Wang, Yuanqing
    Zhang, Doudou
    Liao, Xiangyun
    Cai, Xiaodong
    FRONTIERS IN NEUROLOGY, 2023, 14
  • [2] Characterizing the penumbras of white matter hyperintensities in patients with cerebral small vessel disease
    Xin Wang
    Yu Wang
    Deyu Gao
    Zhichao Zhao
    Haiping Wang
    Sujie Wang
    Shiguang Liu
    Japanese Journal of Radiology, 2023, 41 : 928 - 937
  • [3] Characterizing the penumbras of white matter hyperintensities in patients with cerebral small vessel disease
    Wang, Xin
    Wang, Yu
    Gao, Deyu
    Zhao, Zhichao
    Wang, Haiping
    Wang, Sujie
    Liu, Shiguang
    JAPANESE JOURNAL OF RADIOLOGY, 2023, 41 (09) : 928 - 937
  • [4] Deep-Learning-Based Segmentation and Localization of White Matter Hyperintensities on Magnetic Resonance Images
    Jiang, Wenhao
    Lin, Fengyu
    Zhang, Jian
    Zhan, Taowei
    Cao, Peng
    Wang, Silun
    INTERDISCIPLINARY SCIENCES-COMPUTATIONAL LIFE SCIENCES, 2020, 12 (04) : 438 - 446
  • [5] Deep-Learning-Based Segmentation and Localization of White Matter Hyperintensities on Magnetic Resonance Images
    Wenhao Jiang
    Fengyu Lin
    Jian Zhang
    Taowei Zhan
    Peng Cao
    Silun Wang
    Interdisciplinary Sciences: Computational Life Sciences, 2020, 12 : 438 - 446
  • [6] Performance of antisaccades in patients with cerebral small vessel disease accompanied by white matter hyperintensities
    Fan, Huimin
    Feng, Jing
    Ding, Yuchuan
    Gu, Pan
    Wang, Liying
    Chen, Xiaomeng
    Geng, Xiaokun
    NEUROLOGICAL RESEARCH, 2024, 46 (10) : 917 - 924
  • [7] Association of coagulation markers with the severity of white matter hyperintensities in cerebral small vessel disease
    Xu, Mingyuan
    Li, Jingjing
    Xu, Bu
    Zheng, Qin
    Sun, Wenjun
    FRONTIERS IN NEUROLOGY, 2024, 15
  • [8] Associations between carotid plaques and white matter hyperintensities in cerebral small vessel disease
    Pan, Xijuan
    Liu, Yumei
    Zhou, Fubo
    Tao, Yunlu
    Liu, Ran
    Tian, Bing
    Li, Na
    Chen, Songwei
    Xing, Yingqi
    JOURNAL OF CLINICAL NEUROSCIENCE, 2024, 129
  • [9] HETEROGENEITY OF WHITE MATTER HYPERINTENSITIES IN COGNITIVELY IMPAIRED PATIENTS WITH CEREBRAL SMALL VESSEL DISEASE
    Wang, T.
    Jin, A.
    Fu, Y.
    Zhang, Z.
    Li, S.
    Wang, D.
    Wang, Y.
    INTERNATIONAL JOURNAL OF STROKE, 2023, 18 (03) : 210 - 210
  • [10] Automated detection of white matter hyperintensities of all sizes in cerebral small vessel disease
    Ghafoorian, Mohsen
    Karssemeijer, Nico
    van Uden, Inge W. M.
    de Leeuw, Frank-Erik
    Heskes, Tom
    Marchiori, Elena
    Platel, Bram
    MEDICAL PHYSICS, 2016, 43 (12) : 6246 - 6258