Segmentation Method of Stroke Lesions Based on Bidirectional Recurrent U-Net Model
Stroke has a high disability rate and fatality rate. It is of great clinical significance to study the automatic recognition and segmentation of stroke lesions. Convolutional neural network can not make use of time sequence correlation of medical image data, and has the problem of low feature utiliz...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Editorial Office of Journal of Taiyuan University of Technology
2022-11-01
|
Series: | Taiyuan Ligong Daxue xuebao |
Subjects: | |
Online Access: | https://tyutjournal.tyut.edu.cn/englishpaper/show-1996.html |
_version_ | 1797208328574599168 |
---|---|
author | Xiaohui JIA Xueying ZHANG Suzhe WANG Haisheng HUI Fenglian LI Hua ZHANG |
author_facet | Xiaohui JIA Xueying ZHANG Suzhe WANG Haisheng HUI Fenglian LI Hua ZHANG |
author_sort | Xiaohui JIA |
collection | DOAJ |
description | Stroke has a high disability rate and fatality rate. It is of great clinical significance to study the automatic recognition and segmentation of stroke lesions. Convolutional neural network can not make use of time sequence correlation of medical image data, and has the problem of low feature utilization. Therefore, a bidirectional recurrent U-Net (BIRU-Net) model was proposed for segmentation of lesions. First, a recurrent neural network is introduced to replace part of the convolutional layer in U-Net with an improved attention convolutional gate recursive unit (ACGRU), so that the segmentation model is not only suitable for small-scale annotated medical image data sets, but also has the characteristics of long-term memory. Second, a dual-channel fusion training mechanism is adopted to input the forward and reverse slice data of a single view into BIRU-Net at the same time, and realize bidirectional feature fusion in the process of model forward propagation, effectively utilizing the bidirectional dependence of slice sequence. Finally, the segmentation results of each single view are refused to effectively utilize the spatial context information of data. The experimental results of ATLAS data set show that the DSC value of the proposed method reaches 62.58%. Compared with other methods at the present stage, the proposed method can segment the lesion region more accurately. |
first_indexed | 2024-04-24T09:37:04Z |
format | Article |
id | doaj.art-f6d838e9b55f4a05951c6e3acf85fbda |
institution | Directory Open Access Journal |
issn | 1007-9432 |
language | English |
last_indexed | 2024-04-24T09:37:04Z |
publishDate | 2022-11-01 |
publisher | Editorial Office of Journal of Taiyuan University of Technology |
record_format | Article |
series | Taiyuan Ligong Daxue xuebao |
spelling | doaj.art-f6d838e9b55f4a05951c6e3acf85fbda2024-04-15T09:16:10ZengEditorial Office of Journal of Taiyuan University of TechnologyTaiyuan Ligong Daxue xuebao1007-94322022-11-015361127113310.16355/j.cnki.issn1007-9432tyut.2022.06.0191007-9432(2022)06-1127-07Segmentation Method of Stroke Lesions Based on Bidirectional Recurrent U-Net ModelXiaohui JIA0Xueying ZHANG1Suzhe WANG2Haisheng HUI3Fenglian LI4Hua ZHANG5College of Information and Computer, Taiyuan University of Technology, Taiyuan 030024, ChinaCollege of Information and Computer, Taiyuan University of Technology, Taiyuan 030024, ChinaCollege of Information and Computer, Taiyuan University of Technology, Taiyuan 030024, ChinaCollege of Information and Computer, Taiyuan University of Technology, Taiyuan 030024, ChinaCollege of Information and Computer, Taiyuan University of Technology, Taiyuan 030024, ChinaImaging Department, The First Hospital of Shanxi Medical University, Taiyuan 030024, ChinaStroke has a high disability rate and fatality rate. It is of great clinical significance to study the automatic recognition and segmentation of stroke lesions. Convolutional neural network can not make use of time sequence correlation of medical image data, and has the problem of low feature utilization. Therefore, a bidirectional recurrent U-Net (BIRU-Net) model was proposed for segmentation of lesions. First, a recurrent neural network is introduced to replace part of the convolutional layer in U-Net with an improved attention convolutional gate recursive unit (ACGRU), so that the segmentation model is not only suitable for small-scale annotated medical image data sets, but also has the characteristics of long-term memory. Second, a dual-channel fusion training mechanism is adopted to input the forward and reverse slice data of a single view into BIRU-Net at the same time, and realize bidirectional feature fusion in the process of model forward propagation, effectively utilizing the bidirectional dependence of slice sequence. Finally, the segmentation results of each single view are refused to effectively utilize the spatial context information of data. The experimental results of ATLAS data set show that the DSC value of the proposed method reaches 62.58%. Compared with other methods at the present stage, the proposed method can segment the lesion region more accurately.https://tyutjournal.tyut.edu.cn/englishpaper/show-1996.htmldeep learningsegmentation of stroke lesionscgruu-netbidirectional feature fusionmultiplane fusion |
spellingShingle | Xiaohui JIA Xueying ZHANG Suzhe WANG Haisheng HUI Fenglian LI Hua ZHANG Segmentation Method of Stroke Lesions Based on Bidirectional Recurrent U-Net Model Taiyuan Ligong Daxue xuebao deep learning segmentation of stroke lesions cgru u-net bidirectional feature fusion multiplane fusion |
title | Segmentation Method of Stroke Lesions Based on Bidirectional Recurrent U-Net Model |
title_full | Segmentation Method of Stroke Lesions Based on Bidirectional Recurrent U-Net Model |
title_fullStr | Segmentation Method of Stroke Lesions Based on Bidirectional Recurrent U-Net Model |
title_full_unstemmed | Segmentation Method of Stroke Lesions Based on Bidirectional Recurrent U-Net Model |
title_short | Segmentation Method of Stroke Lesions Based on Bidirectional Recurrent U-Net Model |
title_sort | segmentation method of stroke lesions based on bidirectional recurrent u net model |
topic | deep learning segmentation of stroke lesions cgru u-net bidirectional feature fusion multiplane fusion |
url | https://tyutjournal.tyut.edu.cn/englishpaper/show-1996.html |
work_keys_str_mv | AT xiaohuijia segmentationmethodofstrokelesionsbasedonbidirectionalrecurrentunetmodel AT xueyingzhang segmentationmethodofstrokelesionsbasedonbidirectionalrecurrentunetmodel AT suzhewang segmentationmethodofstrokelesionsbasedonbidirectionalrecurrentunetmodel AT haishenghui segmentationmethodofstrokelesionsbasedonbidirectionalrecurrentunetmodel AT fenglianli segmentationmethodofstrokelesionsbasedonbidirectionalrecurrentunetmodel AT huazhang segmentationmethodofstrokelesionsbasedonbidirectionalrecurrentunetmodel |