Automatic segmentation of rectal tumors from MRI using multiscale densely connected convolutional neural network based on attention mechanism
Rectal cancer is one of the most common malignancies in the gastrointestinal tract. Currently, magnetic resonance imaging has become a vital tool in diagnosing and treating patients with rectal cancer. Notably, early diagnosis of rectal cancer can help improve patient survival rate; however, the cli...
Gespeichert in:
Veröffentlicht in: | Physics in medicine & biology 2023-08, Vol.68 (16), p.165001 |
---|---|
Hauptverfasser: | , , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
container_end_page | |
---|---|
container_issue | 16 |
container_start_page | 165001 |
container_title | Physics in medicine & biology |
container_volume | 68 |
creator | Zhang, Kenan Yang, Xiaotang Cui, Yanfen Zhao, Jumin Li, Dengao |
description | Rectal cancer is one of the most common malignancies in the gastrointestinal tract. Currently, magnetic resonance imaging has become a vital tool in diagnosing and treating patients with rectal cancer. Notably, early diagnosis of rectal cancer can help improve patient survival rate; however, the clinical expertise of physicians is a limiting factor. Therefore, we propose an attention-based multiscale densely connected convolutional neural network based on an attention mechanism to improve the accuracy of diagnosis by automatically segmenting rectal tumors from two-dimensional (2D) magnetic resonance images (MRI) using computer-aided diagnostic techniques. First, to address the inability of U-Net (a classical segmentation network for medical images) and extract rich semantic features and the inconsistent shape and size of tumors between different patients, we replace the conventional convolutional blocks in the U-Net network with multiscale densely connected convolutional blocks. Second, to make the network focus better on global contextual information, we add central blocks with atrous convolution in the final coding layer or the last coding layer. Finally, we add a hybrid attention mechanism to each decoder module to help the model focus on the features of the rectal tumor region. We validated the effectiveness of the proposed method using 3773 2D MRI datasets from 572 patients. The sensitivity, specificity, Dice correlation coefficient, and Hausdorff distance of MRI rectal tumor segmentation were 85.47%, 86.35%, 94.71%, and 7.88mm, respectively. The results showed that the proposed method outperforms conventional approaches. Moreover, the proposed method has better segmentation results in the rectal tumor segmentation task and can provide physicians with the second-most important clinical diagnostic opinion. |
doi_str_mv | 10.1088/1361-6560/ace6f2 |
format | Article |
fullrecord | <record><control><sourceid>proquest_pubme</sourceid><recordid>TN_cdi_pubmed_primary_37437591</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2836876752</sourcerecordid><originalsourceid>FETCH-LOGICAL-c322t-4b8480dffbd2d7f0971ae17d110a236fbc373b1114f1817f68ee15a879fc09653</originalsourceid><addsrcrecordid>eNp9kU9rFjEQxoMo9rV69yQ59uDazGY3yXssRWuhIoieQzY7qVs3mzV_Kv0Q_c5mfWtPIgRmCL95HuYZQl4DewdMqVPgAhrRC3ZqLArXPiG7x6-nZMcYh2YPfX9EXqR0wxiAarvn5IjLjst-Dztyf1Zy8CZPlia89rjk2oeFBkcj2mxmmosPMVEXg6efvlzSkqblmvoy5ylZMyMdcUk431EblqWO4Lh1t2Eum1AVWLDEPyX_CvEHHUyqSLUwOVe7zcyj_W6WKfmX5Jkzc8JXD_WYfPvw_uv5x-bq88Xl-dlVY3nb5qYbVKfY6NwwtqN0bC_BIMgRgJmWCzdYLvkAAJ0DBdIJhQi9UXLvLNuLnh-Tk4PuGsPPgilrX5fBeTYLhpJ0q7hQUsi-rSg7oDaGlCI6vcbJm3ingentCHpLXG-J68MR6sibB_UyeBwfB_6mXoG3B2AKq74JJdaY0v_0Tv6Br37QotKivr6eVq-j478B13Oh5Q</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2836876752</pqid></control><display><type>article</type><title>Automatic segmentation of rectal tumors from MRI using multiscale densely connected convolutional neural network based on attention mechanism</title><source>IOP Publishing Journals</source><source>Institute of Physics (IOP) Journals - HEAL-Link</source><creator>Zhang, Kenan ; Yang, Xiaotang ; Cui, Yanfen ; Zhao, Jumin ; Li, Dengao</creator><creatorcontrib>Zhang, Kenan ; Yang, Xiaotang ; Cui, Yanfen ; Zhao, Jumin ; Li, Dengao</creatorcontrib><description>Rectal cancer is one of the most common malignancies in the gastrointestinal tract. Currently, magnetic resonance imaging has become a vital tool in diagnosing and treating patients with rectal cancer. Notably, early diagnosis of rectal cancer can help improve patient survival rate; however, the clinical expertise of physicians is a limiting factor. Therefore, we propose an attention-based multiscale densely connected convolutional neural network based on an attention mechanism to improve the accuracy of diagnosis by automatically segmenting rectal tumors from two-dimensional (2D) magnetic resonance images (MRI) using computer-aided diagnostic techniques. First, to address the inability of U-Net (a classical segmentation network for medical images) and extract rich semantic features and the inconsistent shape and size of tumors between different patients, we replace the conventional convolutional blocks in the U-Net network with multiscale densely connected convolutional blocks. Second, to make the network focus better on global contextual information, we add central blocks with atrous convolution in the final coding layer or the last coding layer. Finally, we add a hybrid attention mechanism to each decoder module to help the model focus on the features of the rectal tumor region. We validated the effectiveness of the proposed method using 3773 2D MRI datasets from 572 patients. The sensitivity, specificity, Dice correlation coefficient, and Hausdorff distance of MRI rectal tumor segmentation were 85.47%, 86.35%, 94.71%, and 7.88mm, respectively. The results showed that the proposed method outperforms conventional approaches. Moreover, the proposed method has better segmentation results in the rectal tumor segmentation task and can provide physicians with the second-most important clinical diagnostic opinion.</description><identifier>ISSN: 0031-9155</identifier><identifier>EISSN: 1361-6560</identifier><identifier>DOI: 10.1088/1361-6560/ace6f2</identifier><identifier>PMID: 37437591</identifier><identifier>CODEN: PHMBA7</identifier><language>eng</language><publisher>England: IOP Publishing</publisher><subject>attention mechanism ; computer-aided diagnosis ; magnetic resonance imaging ; multiscale dense connectivity ; rectal cancer tumor segmentation</subject><ispartof>Physics in medicine & biology, 2023-08, Vol.68 (16), p.165001</ispartof><rights>2023 Institute of Physics and Engineering in Medicine</rights><rights>2023 Institute of Physics and Engineering in Medicine.</rights><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><cites>FETCH-LOGICAL-c322t-4b8480dffbd2d7f0971ae17d110a236fbc373b1114f1817f68ee15a879fc09653</cites><orcidid>0000-0002-8310-7684</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktopdf>$$Uhttps://iopscience.iop.org/article/10.1088/1361-6560/ace6f2/pdf$$EPDF$$P50$$Giop$$H</linktopdf><link.rule.ids>314,780,784,27924,27925,53846,53893</link.rule.ids><backlink>$$Uhttps://www.ncbi.nlm.nih.gov/pubmed/37437591$$D View this record in MEDLINE/PubMed$$Hfree_for_read</backlink></links><search><creatorcontrib>Zhang, Kenan</creatorcontrib><creatorcontrib>Yang, Xiaotang</creatorcontrib><creatorcontrib>Cui, Yanfen</creatorcontrib><creatorcontrib>Zhao, Jumin</creatorcontrib><creatorcontrib>Li, Dengao</creatorcontrib><title>Automatic segmentation of rectal tumors from MRI using multiscale densely connected convolutional neural network based on attention mechanism</title><title>Physics in medicine & biology</title><addtitle>PMB</addtitle><addtitle>Phys. Med. Biol</addtitle><description>Rectal cancer is one of the most common malignancies in the gastrointestinal tract. Currently, magnetic resonance imaging has become a vital tool in diagnosing and treating patients with rectal cancer. Notably, early diagnosis of rectal cancer can help improve patient survival rate; however, the clinical expertise of physicians is a limiting factor. Therefore, we propose an attention-based multiscale densely connected convolutional neural network based on an attention mechanism to improve the accuracy of diagnosis by automatically segmenting rectal tumors from two-dimensional (2D) magnetic resonance images (MRI) using computer-aided diagnostic techniques. First, to address the inability of U-Net (a classical segmentation network for medical images) and extract rich semantic features and the inconsistent shape and size of tumors between different patients, we replace the conventional convolutional blocks in the U-Net network with multiscale densely connected convolutional blocks. Second, to make the network focus better on global contextual information, we add central blocks with atrous convolution in the final coding layer or the last coding layer. Finally, we add a hybrid attention mechanism to each decoder module to help the model focus on the features of the rectal tumor region. We validated the effectiveness of the proposed method using 3773 2D MRI datasets from 572 patients. The sensitivity, specificity, Dice correlation coefficient, and Hausdorff distance of MRI rectal tumor segmentation were 85.47%, 86.35%, 94.71%, and 7.88mm, respectively. The results showed that the proposed method outperforms conventional approaches. Moreover, the proposed method has better segmentation results in the rectal tumor segmentation task and can provide physicians with the second-most important clinical diagnostic opinion.</description><subject>attention mechanism</subject><subject>computer-aided diagnosis</subject><subject>magnetic resonance imaging</subject><subject>multiscale dense connectivity</subject><subject>rectal cancer tumor segmentation</subject><issn>0031-9155</issn><issn>1361-6560</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><recordid>eNp9kU9rFjEQxoMo9rV69yQ59uDazGY3yXssRWuhIoieQzY7qVs3mzV_Kv0Q_c5mfWtPIgRmCL95HuYZQl4DewdMqVPgAhrRC3ZqLArXPiG7x6-nZMcYh2YPfX9EXqR0wxiAarvn5IjLjst-Dztyf1Zy8CZPlia89rjk2oeFBkcj2mxmmosPMVEXg6efvlzSkqblmvoy5ylZMyMdcUk431EblqWO4Lh1t2Eum1AVWLDEPyX_CvEHHUyqSLUwOVe7zcyj_W6WKfmX5Jkzc8JXD_WYfPvw_uv5x-bq88Xl-dlVY3nb5qYbVKfY6NwwtqN0bC_BIMgRgJmWCzdYLvkAAJ0DBdIJhQi9UXLvLNuLnh-Tk4PuGsPPgilrX5fBeTYLhpJ0q7hQUsi-rSg7oDaGlCI6vcbJm3ingentCHpLXG-J68MR6sibB_UyeBwfB_6mXoG3B2AKq74JJdaY0v_0Tv6Br37QotKivr6eVq-j478B13Oh5Q</recordid><startdate>20230821</startdate><enddate>20230821</enddate><creator>Zhang, Kenan</creator><creator>Yang, Xiaotang</creator><creator>Cui, Yanfen</creator><creator>Zhao, Jumin</creator><creator>Li, Dengao</creator><general>IOP Publishing</general><scope>NPM</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7X8</scope><orcidid>https://orcid.org/0000-0002-8310-7684</orcidid></search><sort><creationdate>20230821</creationdate><title>Automatic segmentation of rectal tumors from MRI using multiscale densely connected convolutional neural network based on attention mechanism</title><author>Zhang, Kenan ; Yang, Xiaotang ; Cui, Yanfen ; Zhao, Jumin ; Li, Dengao</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c322t-4b8480dffbd2d7f0971ae17d110a236fbc373b1114f1817f68ee15a879fc09653</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><topic>attention mechanism</topic><topic>computer-aided diagnosis</topic><topic>magnetic resonance imaging</topic><topic>multiscale dense connectivity</topic><topic>rectal cancer tumor segmentation</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Zhang, Kenan</creatorcontrib><creatorcontrib>Yang, Xiaotang</creatorcontrib><creatorcontrib>Cui, Yanfen</creatorcontrib><creatorcontrib>Zhao, Jumin</creatorcontrib><creatorcontrib>Li, Dengao</creatorcontrib><collection>PubMed</collection><collection>CrossRef</collection><collection>MEDLINE - Academic</collection><jtitle>Physics in medicine & biology</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Zhang, Kenan</au><au>Yang, Xiaotang</au><au>Cui, Yanfen</au><au>Zhao, Jumin</au><au>Li, Dengao</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Automatic segmentation of rectal tumors from MRI using multiscale densely connected convolutional neural network based on attention mechanism</atitle><jtitle>Physics in medicine & biology</jtitle><stitle>PMB</stitle><addtitle>Phys. Med. Biol</addtitle><date>2023-08-21</date><risdate>2023</risdate><volume>68</volume><issue>16</issue><spage>165001</spage><pages>165001-</pages><issn>0031-9155</issn><eissn>1361-6560</eissn><coden>PHMBA7</coden><abstract>Rectal cancer is one of the most common malignancies in the gastrointestinal tract. Currently, magnetic resonance imaging has become a vital tool in diagnosing and treating patients with rectal cancer. Notably, early diagnosis of rectal cancer can help improve patient survival rate; however, the clinical expertise of physicians is a limiting factor. Therefore, we propose an attention-based multiscale densely connected convolutional neural network based on an attention mechanism to improve the accuracy of diagnosis by automatically segmenting rectal tumors from two-dimensional (2D) magnetic resonance images (MRI) using computer-aided diagnostic techniques. First, to address the inability of U-Net (a classical segmentation network for medical images) and extract rich semantic features and the inconsistent shape and size of tumors between different patients, we replace the conventional convolutional blocks in the U-Net network with multiscale densely connected convolutional blocks. Second, to make the network focus better on global contextual information, we add central blocks with atrous convolution in the final coding layer or the last coding layer. Finally, we add a hybrid attention mechanism to each decoder module to help the model focus on the features of the rectal tumor region. We validated the effectiveness of the proposed method using 3773 2D MRI datasets from 572 patients. The sensitivity, specificity, Dice correlation coefficient, and Hausdorff distance of MRI rectal tumor segmentation were 85.47%, 86.35%, 94.71%, and 7.88mm, respectively. The results showed that the proposed method outperforms conventional approaches. Moreover, the proposed method has better segmentation results in the rectal tumor segmentation task and can provide physicians with the second-most important clinical diagnostic opinion.</abstract><cop>England</cop><pub>IOP Publishing</pub><pmid>37437591</pmid><doi>10.1088/1361-6560/ace6f2</doi><tpages>16</tpages><orcidid>https://orcid.org/0000-0002-8310-7684</orcidid></addata></record> |
fulltext | fulltext |
identifier | ISSN: 0031-9155 |
ispartof | Physics in medicine & biology, 2023-08, Vol.68 (16), p.165001 |
issn | 0031-9155 1361-6560 |
language | eng |
recordid | cdi_pubmed_primary_37437591 |
source | IOP Publishing Journals; Institute of Physics (IOP) Journals - HEAL-Link |
subjects | attention mechanism computer-aided diagnosis magnetic resonance imaging multiscale dense connectivity rectal cancer tumor segmentation |
title | Automatic segmentation of rectal tumors from MRI using multiscale densely connected convolutional neural network based on attention mechanism |
url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-25T18%3A32%3A45IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_pubme&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Automatic%20segmentation%20of%20rectal%20tumors%20from%20MRI%20using%20multiscale%20densely%20connected%20convolutional%20neural%20network%20based%20on%20attention%20mechanism&rft.jtitle=Physics%20in%20medicine%20&%20biology&rft.au=Zhang,%20Kenan&rft.date=2023-08-21&rft.volume=68&rft.issue=16&rft.spage=165001&rft.pages=165001-&rft.issn=0031-9155&rft.eissn=1361-6560&rft.coden=PHMBA7&rft_id=info:doi/10.1088/1361-6560/ace6f2&rft_dat=%3Cproquest_pubme%3E2836876752%3C/proquest_pubme%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2836876752&rft_id=info:pmid/37437591&rfr_iscdi=true |