Cargando…
SeUneter: Channel attentive U-Net for instance segmentation of the cervical spine MRI medical image
In recent years, cervical spondylosis has become one of the most common chronic diseases and has received much attention from the public. Magnetic resonance imaging (MRI) is the most widely used imaging modality for the diagnosis of degenerative cervical spondylosis. The manual identification and se...
Autores principales: | , , , , , , , , , , , , |
---|---|
Formato: | Online Artículo Texto |
Lenguaje: | English |
Publicado: |
Frontiers Media S.A.
2022
|
Materias: | |
Acceso en línea: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9763568/ https://www.ncbi.nlm.nih.gov/pubmed/36561215 http://dx.doi.org/10.3389/fphys.2022.1081441 |
_version_ | 1784853087575343104 |
---|---|
author | Zhang, Xiang Yang, Yi Shen, Yi-Wei Li, Ping Zhong, Yuan Zhou, Jing Zhang, Ke-Rui Shen, Chang-Yong Li, Yi Zhang, Meng-Fei Pan, Long-Hai Ma, Li-Tai Liu, Hao |
author_facet | Zhang, Xiang Yang, Yi Shen, Yi-Wei Li, Ping Zhong, Yuan Zhou, Jing Zhang, Ke-Rui Shen, Chang-Yong Li, Yi Zhang, Meng-Fei Pan, Long-Hai Ma, Li-Tai Liu, Hao |
author_sort | Zhang, Xiang |
collection | PubMed |
description | In recent years, cervical spondylosis has become one of the most common chronic diseases and has received much attention from the public. Magnetic resonance imaging (MRI) is the most widely used imaging modality for the diagnosis of degenerative cervical spondylosis. The manual identification and segmentation of the cervical spine on MRI makes it a laborious, time-consuming, and error-prone process. In this work, we collected a new dataset of 300 patients with a total of 600 cervical spine images in the MRI T2-weighted (T2W) modality for the first time, which included the cervical spine, intervertebral discs, spinal cord, and spinal canal information. A new instance segmentation approach called SeUneter was proposed for cervical spine segmentation. SeUneter expanded the depth of the network structure based on the original U-Net and added a channel attention module to the double convolution of the feature extraction. SeUneter could enhance the semantic information of the segmentation and weaken the characteristic information of non-segmentation to the screen for important feature channels in double convolution. In the meantime, to alleviate the over-fitting of the model under insufficient samples, the Cutout was used to crop the pixel information in the original image at random positions of a fixed size, and the number of training samples in the original data was increased. Prior knowledge of the data was used to optimize the segmentation results by a post-process to improve the segmentation performance. The mean of Intersection Over Union (mIOU) was calculated for the different categories, while the mean of the Dice similarity coefficient (mDSC) and mIOU were calculated to compare the segmentation results of different deep learning models for all categories. Compared with multiple models under the same experimental settings, our proposed SeUneter’s performance was superior to U-Net, AttU-Net, UNet++, DeepLab-v3+, TransUNet, and Swin-Unet on the spinal cord with mIOU of 86.34% and the spinal canal with mIOU of 73.44%. The SeUneter matched or exceeded the performance of the aforementioned segmentation models when segmenting vertebral bodies or intervertebral discs. Among all models, SeUneter achieved the highest mIOU and mDSC of 82.73% and 90.66%, respectively, for the whole cervical spine. |
format | Online Article Text |
id | pubmed-9763568 |
institution | National Center for Biotechnology Information |
language | English |
publishDate | 2022 |
publisher | Frontiers Media S.A. |
record_format | MEDLINE/PubMed |
spelling | pubmed-97635682022-12-21 SeUneter: Channel attentive U-Net for instance segmentation of the cervical spine MRI medical image Zhang, Xiang Yang, Yi Shen, Yi-Wei Li, Ping Zhong, Yuan Zhou, Jing Zhang, Ke-Rui Shen, Chang-Yong Li, Yi Zhang, Meng-Fei Pan, Long-Hai Ma, Li-Tai Liu, Hao Front Physiol Physiology In recent years, cervical spondylosis has become one of the most common chronic diseases and has received much attention from the public. Magnetic resonance imaging (MRI) is the most widely used imaging modality for the diagnosis of degenerative cervical spondylosis. The manual identification and segmentation of the cervical spine on MRI makes it a laborious, time-consuming, and error-prone process. In this work, we collected a new dataset of 300 patients with a total of 600 cervical spine images in the MRI T2-weighted (T2W) modality for the first time, which included the cervical spine, intervertebral discs, spinal cord, and spinal canal information. A new instance segmentation approach called SeUneter was proposed for cervical spine segmentation. SeUneter expanded the depth of the network structure based on the original U-Net and added a channel attention module to the double convolution of the feature extraction. SeUneter could enhance the semantic information of the segmentation and weaken the characteristic information of non-segmentation to the screen for important feature channels in double convolution. In the meantime, to alleviate the over-fitting of the model under insufficient samples, the Cutout was used to crop the pixel information in the original image at random positions of a fixed size, and the number of training samples in the original data was increased. Prior knowledge of the data was used to optimize the segmentation results by a post-process to improve the segmentation performance. The mean of Intersection Over Union (mIOU) was calculated for the different categories, while the mean of the Dice similarity coefficient (mDSC) and mIOU were calculated to compare the segmentation results of different deep learning models for all categories. Compared with multiple models under the same experimental settings, our proposed SeUneter’s performance was superior to U-Net, AttU-Net, UNet++, DeepLab-v3+, TransUNet, and Swin-Unet on the spinal cord with mIOU of 86.34% and the spinal canal with mIOU of 73.44%. The SeUneter matched or exceeded the performance of the aforementioned segmentation models when segmenting vertebral bodies or intervertebral discs. Among all models, SeUneter achieved the highest mIOU and mDSC of 82.73% and 90.66%, respectively, for the whole cervical spine. Frontiers Media S.A. 2022-12-06 /pmc/articles/PMC9763568/ /pubmed/36561215 http://dx.doi.org/10.3389/fphys.2022.1081441 Text en Copyright © 2022 Zhang, Yang, Shen, Li, Zhong, Zhou, Zhang, Shen, Li, Zhang, Pan, Ma and Liu. https://creativecommons.org/licenses/by/4.0/This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. |
spellingShingle | Physiology Zhang, Xiang Yang, Yi Shen, Yi-Wei Li, Ping Zhong, Yuan Zhou, Jing Zhang, Ke-Rui Shen, Chang-Yong Li, Yi Zhang, Meng-Fei Pan, Long-Hai Ma, Li-Tai Liu, Hao SeUneter: Channel attentive U-Net for instance segmentation of the cervical spine MRI medical image |
title | SeUneter: Channel attentive U-Net for instance segmentation of the cervical spine MRI medical image |
title_full | SeUneter: Channel attentive U-Net for instance segmentation of the cervical spine MRI medical image |
title_fullStr | SeUneter: Channel attentive U-Net for instance segmentation of the cervical spine MRI medical image |
title_full_unstemmed | SeUneter: Channel attentive U-Net for instance segmentation of the cervical spine MRI medical image |
title_short | SeUneter: Channel attentive U-Net for instance segmentation of the cervical spine MRI medical image |
title_sort | seuneter: channel attentive u-net for instance segmentation of the cervical spine mri medical image |
topic | Physiology |
url | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9763568/ https://www.ncbi.nlm.nih.gov/pubmed/36561215 http://dx.doi.org/10.3389/fphys.2022.1081441 |
work_keys_str_mv | AT zhangxiang seuneterchannelattentiveunetforinstancesegmentationofthecervicalspinemrimedicalimage AT yangyi seuneterchannelattentiveunetforinstancesegmentationofthecervicalspinemrimedicalimage AT shenyiwei seuneterchannelattentiveunetforinstancesegmentationofthecervicalspinemrimedicalimage AT liping seuneterchannelattentiveunetforinstancesegmentationofthecervicalspinemrimedicalimage AT zhongyuan seuneterchannelattentiveunetforinstancesegmentationofthecervicalspinemrimedicalimage AT zhoujing seuneterchannelattentiveunetforinstancesegmentationofthecervicalspinemrimedicalimage AT zhangkerui seuneterchannelattentiveunetforinstancesegmentationofthecervicalspinemrimedicalimage AT shenchangyong seuneterchannelattentiveunetforinstancesegmentationofthecervicalspinemrimedicalimage AT liyi seuneterchannelattentiveunetforinstancesegmentationofthecervicalspinemrimedicalimage AT zhangmengfei seuneterchannelattentiveunetforinstancesegmentationofthecervicalspinemrimedicalimage AT panlonghai seuneterchannelattentiveunetforinstancesegmentationofthecervicalspinemrimedicalimage AT malitai seuneterchannelattentiveunetforinstancesegmentationofthecervicalspinemrimedicalimage AT liuhao seuneterchannelattentiveunetforinstancesegmentationofthecervicalspinemrimedicalimage |