File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Article: CT-based deep learning segmentation of ovarian cancer and the stability of the extracted radiomics features

TitleCT-based deep learning segmentation of ovarian cancer and the stability of the extracted radiomics features
Authors
Keywordsautomated segmentation
computed tomography
deep learning
Ovarian cancer (OC)
radiomics
Issue Date13-Jun-2023
PublisherAME Publishing
Citation
Quantitative Imaging in Medicine and Surgery, 2023, v. 13, n. 8, p. 5218-5229 How to Cite?
Abstract

Background: Radiomics analysis could provide complementary tissue characterization in ovarian cancer (OC). However, OC segmentation required in radiomics analysis is time-consuming and labour-intensive. In this study, we aim to evaluate the performance of deep learning-based segmentation of OC on contrast-enhanced CT images and the stability of radiomics features extracted from the automated segmentation.

Methods: Staging abdominopelvic CT images of 367 patients with OC were retrospectively recruited. The training and cross-validation sets came from center A (n=283), and testing set (n=84) came from centers B and C. The tumours were manually delineated by a board-certified radiologist. Four model architectures provided by no-new-Net (nnU-Net) method were tested in this task. The segmentation performance evaluated by Dice score, Jaccard score, sensitivity and precision were compared among 4 architectures. The Pearson correlation coefficient (ρ), concordance correlation coefficient (ρc) and Bland-Altman plots were used to evaluate the volumetric assessment of OC between manual and automated segmentations. The stability of extracted radiomics features was evaluated by intraclass correlation coefficient (ICC).

Results: The 3D U-Net cascade architecture achieved highest median Dice score, Jaccard score, sensitivity and precision for OC segmentation in the testing set, 0.941, 0.890, 0.973 and 0.925, respectively. Tumour volumes of manual and automated segmentations were highly correlated (ρ=0.944 and ρc =0.933). 85.0% of radiomics features had high correlation with ICC >0.8.

Conclusions: The presented deep-learning segmentation could provide highly accurate automated segmentation of OC on CT images with high stability of the extracted radiomics features, showing the potential as a batch-processing segmentation tool.


Persistent Identifierhttp://hdl.handle.net/10722/331427
ISSN
2023 Impact Factor: 2.9
2023 SCImago Journal Rankings: 0.746
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorWang, Y-
dc.contributor.authorWang, MD-
dc.contributor.authorCao, P-
dc.contributor.authorWong, EMF-
dc.contributor.authorHo, GC-
dc.contributor.authorLam, TPW-
dc.contributor.authorHan, LJ-
dc.contributor.authorLee, EYP-
dc.date.accessioned2023-09-21T06:55:37Z-
dc.date.available2023-09-21T06:55:37Z-
dc.date.issued2023-06-13-
dc.identifier.citationQuantitative Imaging in Medicine and Surgery, 2023, v. 13, n. 8, p. 5218-5229-
dc.identifier.issn2223-4292-
dc.identifier.urihttp://hdl.handle.net/10722/331427-
dc.description.abstract<p><strong>Background: </strong>Radiomics analysis could provide complementary tissue characterization in ovarian cancer (OC). However, OC segmentation required in radiomics analysis is time-consuming and labour-intensive. In this study, we aim to evaluate the performance of deep learning-based segmentation of OC on contrast-enhanced CT images and the stability of radiomics features extracted from the automated segmentation.</p><p><strong>Methods: </strong>Staging abdominopelvic CT images of 367 patients with OC were retrospectively recruited. The training and cross-validation sets came from center A (n=283), and testing set (n=84) came from centers B and C. The tumours were manually delineated by a board-certified radiologist. Four model architectures provided by no-new-Net (nnU-Net) method were tested in this task. The segmentation performance evaluated by Dice score, Jaccard score, sensitivity and precision were compared among 4 architectures. The Pearson correlation coefficient (<em>ρ</em>), concordance correlation coefficient (<em>ρ<sub>c</sub></em>) and Bland-Altman plots were used to evaluate the volumetric assessment of OC between manual and automated segmentations. The stability of extracted radiomics features was evaluated by intraclass correlation coefficient (ICC).</p><p><strong>Results: </strong>The 3D U-Net cascade architecture achieved highest median Dice score, Jaccard score, sensitivity and precision for OC segmentation in the testing set, 0.941, 0.890, 0.973 and 0.925, respectively. Tumour volumes of manual and automated segmentations were highly correlated (<em>ρ</em>=0.944 and <em>ρ<sub>c</sub></em> =0.933). 85.0% of radiomics features had high correlation with ICC >0.8.</p><p><strong>Conclusions: </strong>The presented deep-learning segmentation could provide highly accurate automated segmentation of OC on CT images with high stability of the extracted radiomics features, showing the potential as a batch-processing segmentation tool.</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.subjectautomated segmentation-
dc.subjectcomputed tomography-
dc.subjectdeep learning-
dc.subjectOvarian cancer (OC)-
dc.subjectradiomics-
dc.titleCT-based deep learning segmentation of ovarian cancer and the stability of the extracted radiomics features-
dc.typeArticle-
dc.identifier.doi10.21037/qims-22-1135-
dc.identifier.scopuseid_2-s2.0-85168832757-
dc.identifier.volume13-
dc.identifier.issue8-
dc.identifier.spage5218-
dc.identifier.epage5229-
dc.identifier.eissn2223-4306-
dc.identifier.isiWOS:001009890100001-
dc.identifier.issnl2223-4306-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats