A deep learning algorithm for automatic detection and classification of acute intracranial hemorrhages in head CT scans
Acute Intracranial hemorrhage (ICH) is a life-threatening disease that requires emergency medical attention, which is routinely diagnosed using non-contrast head CT imaging. The diagnostic accuracy of acute ICH on CT varies greatly among radiologists due to the difficulty of interpreting subtle find...
Main Authors: | , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Elsevier
2021-01-01
|
Series: | NeuroImage: Clinical |
Subjects: | |
Online Access: | http://www.sciencedirect.com/science/article/pii/S2213158221002291 |
id |
doaj-f53eec8f558c42e988af383e2bae8045 |
---|---|
record_format |
Article |
spelling |
doaj-f53eec8f558c42e988af383e2bae80452021-08-18T04:22:04ZengElsevierNeuroImage: Clinical2213-15822021-01-0132102785A deep learning algorithm for automatic detection and classification of acute intracranial hemorrhages in head CT scansXiyue Wang0Tao Shen1Sen Yang2Jun Lan3Yanming Xu4Minghui Wang5Jing Zhang6Xiao Han7College of Computer Science, Sichuan University, Chengdu 610065, ChinaTencent AI Lab, Shenzhen 518057, ChinaTencent AI Lab, Shenzhen 518057, ChinaWinning Health Technology Group Co., Ltd, Shanghai, ChinaDepartment of Neurology, West China Hospital, Sichuan University, Chengdu, ChinaCollege of Computer Science, Sichuan University, Chengdu 610065, ChinaCollege of Biomedical Engineering, Sichuan University, Chengdu, China; Corresponding authors.Tencent AI Lab, Shenzhen 518057, China; Corresponding authors.Acute Intracranial hemorrhage (ICH) is a life-threatening disease that requires emergency medical attention, which is routinely diagnosed using non-contrast head CT imaging. The diagnostic accuracy of acute ICH on CT varies greatly among radiologists due to the difficulty of interpreting subtle findings and the time pressure associated with the ever-increasing workload. The use of artificial intelligence technology may help automate the process and assist radiologists for more prompt and better decision-making. In this work, we design a deep learning approach that mimics the interpretation process of radiologists, and combines a 2D CNN model and two sequence models to achieve accurate acute ICH detection and subtype classification. Being developed using the extensive 2019-RSNA Brain CT Hemorrhage Challenge dataset with over 25000 CT scans, our deep learning algorithm can accurately classify the acute ICH and its five subtypes with AUCs of 0.988 (ICH), 0.984 (EDH), 0.992 (IPH), 0.996 (IVH), 0.985 (SAH), and 0.983 (SDH), respectively, reaching the accuracy level of expert radiologists. Our method won 1st place among 1345 teams from 75 countries in the RSNA challenge. We have further evaluated our algorithm on two independent external validation datasets with 75 and 491 CT scans, respectively, and our method maintained high AUCs of 0.964 and 0.949 for acute ICH detection. These results have demonstrated the high performance and robust generalization ability of our proposed method, which makes it a useful second-read or triage tool that can facilitate routine clinical applications.http://www.sciencedirect.com/science/article/pii/S2213158221002291Intracranial hemorrhage (ICH)Head CTDeep learningImage classificationSequence model |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Xiyue Wang Tao Shen Sen Yang Jun Lan Yanming Xu Minghui Wang Jing Zhang Xiao Han |
spellingShingle |
Xiyue Wang Tao Shen Sen Yang Jun Lan Yanming Xu Minghui Wang Jing Zhang Xiao Han A deep learning algorithm for automatic detection and classification of acute intracranial hemorrhages in head CT scans NeuroImage: Clinical Intracranial hemorrhage (ICH) Head CT Deep learning Image classification Sequence model |
author_facet |
Xiyue Wang Tao Shen Sen Yang Jun Lan Yanming Xu Minghui Wang Jing Zhang Xiao Han |
author_sort |
Xiyue Wang |
title |
A deep learning algorithm for automatic detection and classification of acute intracranial hemorrhages in head CT scans |
title_short |
A deep learning algorithm for automatic detection and classification of acute intracranial hemorrhages in head CT scans |
title_full |
A deep learning algorithm for automatic detection and classification of acute intracranial hemorrhages in head CT scans |
title_fullStr |
A deep learning algorithm for automatic detection and classification of acute intracranial hemorrhages in head CT scans |
title_full_unstemmed |
A deep learning algorithm for automatic detection and classification of acute intracranial hemorrhages in head CT scans |
title_sort |
deep learning algorithm for automatic detection and classification of acute intracranial hemorrhages in head ct scans |
publisher |
Elsevier |
series |
NeuroImage: Clinical |
issn |
2213-1582 |
publishDate |
2021-01-01 |
description |
Acute Intracranial hemorrhage (ICH) is a life-threatening disease that requires emergency medical attention, which is routinely diagnosed using non-contrast head CT imaging. The diagnostic accuracy of acute ICH on CT varies greatly among radiologists due to the difficulty of interpreting subtle findings and the time pressure associated with the ever-increasing workload. The use of artificial intelligence technology may help automate the process and assist radiologists for more prompt and better decision-making. In this work, we design a deep learning approach that mimics the interpretation process of radiologists, and combines a 2D CNN model and two sequence models to achieve accurate acute ICH detection and subtype classification. Being developed using the extensive 2019-RSNA Brain CT Hemorrhage Challenge dataset with over 25000 CT scans, our deep learning algorithm can accurately classify the acute ICH and its five subtypes with AUCs of 0.988 (ICH), 0.984 (EDH), 0.992 (IPH), 0.996 (IVH), 0.985 (SAH), and 0.983 (SDH), respectively, reaching the accuracy level of expert radiologists. Our method won 1st place among 1345 teams from 75 countries in the RSNA challenge. We have further evaluated our algorithm on two independent external validation datasets with 75 and 491 CT scans, respectively, and our method maintained high AUCs of 0.964 and 0.949 for acute ICH detection. These results have demonstrated the high performance and robust generalization ability of our proposed method, which makes it a useful second-read or triage tool that can facilitate routine clinical applications. |
topic |
Intracranial hemorrhage (ICH) Head CT Deep learning Image classification Sequence model |
url |
http://www.sciencedirect.com/science/article/pii/S2213158221002291 |
work_keys_str_mv |
AT xiyuewang adeeplearningalgorithmforautomaticdetectionandclassificationofacuteintracranialhemorrhagesinheadctscans AT taoshen adeeplearningalgorithmforautomaticdetectionandclassificationofacuteintracranialhemorrhagesinheadctscans AT senyang adeeplearningalgorithmforautomaticdetectionandclassificationofacuteintracranialhemorrhagesinheadctscans AT junlan adeeplearningalgorithmforautomaticdetectionandclassificationofacuteintracranialhemorrhagesinheadctscans AT yanmingxu adeeplearningalgorithmforautomaticdetectionandclassificationofacuteintracranialhemorrhagesinheadctscans AT minghuiwang adeeplearningalgorithmforautomaticdetectionandclassificationofacuteintracranialhemorrhagesinheadctscans AT jingzhang adeeplearningalgorithmforautomaticdetectionandclassificationofacuteintracranialhemorrhagesinheadctscans AT xiaohan adeeplearningalgorithmforautomaticdetectionandclassificationofacuteintracranialhemorrhagesinheadctscans AT xiyuewang deeplearningalgorithmforautomaticdetectionandclassificationofacuteintracranialhemorrhagesinheadctscans AT taoshen deeplearningalgorithmforautomaticdetectionandclassificationofacuteintracranialhemorrhagesinheadctscans AT senyang deeplearningalgorithmforautomaticdetectionandclassificationofacuteintracranialhemorrhagesinheadctscans AT junlan deeplearningalgorithmforautomaticdetectionandclassificationofacuteintracranialhemorrhagesinheadctscans AT yanmingxu deeplearningalgorithmforautomaticdetectionandclassificationofacuteintracranialhemorrhagesinheadctscans AT minghuiwang deeplearningalgorithmforautomaticdetectionandclassificationofacuteintracranialhemorrhagesinheadctscans AT jingzhang deeplearningalgorithmforautomaticdetectionandclassificationofacuteintracranialhemorrhagesinheadctscans AT xiaohan deeplearningalgorithmforautomaticdetectionandclassificationofacuteintracranialhemorrhagesinheadctscans |
_version_ |
1721203721130475520 |