File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Article: A deep learning model based on the attention mechanism for automatic segmentation of abdominal muscle and fat for body composition assessment

TitleA deep learning model based on the attention mechanism for automatic segmentation of abdominal muscle and fat for body composition assessment
Authors
KeywordsCT scan
deep learning
fat segmentation
muscle segmentation
sarcopenia
Issue Date9-Feb-2023
PublisherAME Publishing
Citation
Quantitative Imaging in Medicine and Surgery, 2023, v. 13, n. 3, p. 1384-1398 How to Cite?
Abstract

Background: Quantitative muscle and fat data obtained through body composition analysis are expected to be a new stable biomarker for the early and accurate prediction of treatment-related toxicity, treatment response, and prognosis in patients with lung cancer. The use of these biomarkers can enable the adjustment of individualized treatment regimens in a timely manner, which is critical to further improving patient prognosis and quality of life. We aimed to develop a deep learning model based on attention for fully automated segmentation of the abdomen from computed tomography (CT) to quantify body composition.

Methods: A fully automatic segmentation deep learning model was designed based on the attention mechanism and using U-Net as the framework. Subcutaneous fat, skeletal muscle, and visceral fat were manually segmented by two experts to serve as ground truth labels. The performance of the model was evaluated using Dice similarity coefficients (DSCs) and Hausdorff distance at 95th percentile (HD95).

Results: The mean DSC for subcutaneous fat and skeletal muscle were high for both the enhanced CT test set (0.93±0.06 and 0.96±0.02, respectively) and the plain CT test set (0.90±0.09 and 0.95±0.01, respectively). Nevertheless, the model did not perform well in the segmentation performance of visceral fat, especially for the enhanced CT test set. The mean DSC for the enhanced CT test set was 0.87±0.11, while the mean DSC for the plain CT test set was 0.92±0.03. We discuss the reasons for this result.

Conclusions: This work demonstrates a method for the automatic outlining of subcutaneous fat, skeletal muscle, and visceral fat areas at L3.


Persistent Identifierhttp://hdl.handle.net/10722/331830
ISSN
2023 Impact Factor: 2.9
2023 SCImago Journal Rankings: 0.746
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorShen, H-
dc.contributor.authorHe, P-
dc.contributor.authorRen, Y-
dc.contributor.authorHuang, ZY-
dc.contributor.authorLi, SL-
dc.contributor.authorWang, GS-
dc.contributor.authorCong, MH-
dc.contributor.authorLuo, DH-
dc.contributor.authorShao, D-
dc.contributor.authorLee, EYP-
dc.contributor.authorCui, RX-
dc.contributor.authorHuo, L-
dc.contributor.authorQin, J-
dc.contributor.authorLiu, J-
dc.contributor.authorHu, ZL-
dc.contributor.authorLiu, Z-
dc.contributor.authorZhang, N-
dc.date.accessioned2023-09-21T06:59:17Z-
dc.date.available2023-09-21T06:59:17Z-
dc.date.issued2023-02-09-
dc.identifier.citationQuantitative Imaging in Medicine and Surgery, 2023, v. 13, n. 3, p. 1384-1398-
dc.identifier.issn2223-4292-
dc.identifier.urihttp://hdl.handle.net/10722/331830-
dc.description.abstract<p><strong>Background: </strong>Quantitative muscle and fat data obtained through body composition analysis are expected to be a new stable biomarker for the early and accurate prediction of treatment-related toxicity, treatment response, and prognosis in patients with lung cancer. The use of these biomarkers can enable the adjustment of individualized treatment regimens in a timely manner, which is critical to further improving patient prognosis and quality of life. We aimed to develop a deep learning model based on attention for fully automated segmentation of the abdomen from computed tomography (CT) to quantify body composition.</p><p><strong>Methods: </strong>A fully automatic segmentation deep learning model was designed based on the attention mechanism and using U-Net as the framework. Subcutaneous fat, skeletal muscle, and visceral fat were manually segmented by two experts to serve as ground truth labels. The performance of the model was evaluated using Dice similarity coefficients (DSCs) and Hausdorff distance at 95th percentile (HD95).</p><p><strong>Results: </strong>The mean DSC for subcutaneous fat and skeletal muscle were high for both the enhanced CT test set (0.93±0.06 and 0.96±0.02, respectively) and the plain CT test set (0.90±0.09 and 0.95±0.01, respectively). Nevertheless, the model did not perform well in the segmentation performance of visceral fat, especially for the enhanced CT test set. The mean DSC for the enhanced CT test set was 0.87±0.11, while the mean DSC for the plain CT test set was 0.92±0.03. We discuss the reasons for this result.</p><p><strong>Conclusions: </strong>This work demonstrates a method for the automatic outlining of subcutaneous fat, skeletal muscle, and visceral fat areas at L3.</p>-
dc.languageeng-
dc.publisherAME Publishing-
dc.relation.ispartofQuantitative Imaging in Medicine and Surgery-
dc.rightsThis work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.-
dc.subjectCT scan-
dc.subjectdeep learning-
dc.subjectfat segmentation-
dc.subjectmuscle segmentation-
dc.subjectsarcopenia-
dc.titleA deep learning model based on the attention mechanism for automatic segmentation of abdominal muscle and fat for body composition assessment-
dc.typeArticle-
dc.identifier.doi10.21037/qims-22-330-
dc.identifier.scopuseid_2-s2.0-85149001593-
dc.identifier.volume13-
dc.identifier.issue3-
dc.identifier.spage1384-
dc.identifier.epage1398-
dc.identifier.eissn2223-4306-
dc.identifier.isiWOS:000992707600002-
dc.identifier.issnl2223-4306-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats