Deep Learning for Automated Elective Lymph Node Level Segmentation for Head and Neck Cancer Radiotherapy
Depending on the clinical situation, different combinations of lymph node (LN) levels define the elective LN target volume in head-and-neck cancer (HNC) radiotherapy. The accurate auto-contouring of individual LN levels could reduce the burden and variability of manual segmentation and be used regar...
Main Authors: | , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2022-11-01
|
Series: | Cancers |
Subjects: | |
Online Access: | https://www.mdpi.com/2072-6694/14/22/5501 |
_version_ | 1797465786307051520 |
---|---|
author | Victor I. J. Strijbis Max Dahele Oliver J. Gurney-Champion Gerrit J. Blom Marije R. Vergeer Berend J. Slotman Wilko F. A. R. Verbakel |
author_facet | Victor I. J. Strijbis Max Dahele Oliver J. Gurney-Champion Gerrit J. Blom Marije R. Vergeer Berend J. Slotman Wilko F. A. R. Verbakel |
author_sort | Victor I. J. Strijbis |
collection | DOAJ |
description | Depending on the clinical situation, different combinations of lymph node (LN) levels define the elective LN target volume in head-and-neck cancer (HNC) radiotherapy. The accurate auto-contouring of individual LN levels could reduce the burden and variability of manual segmentation and be used regardless of the primary tumor location. We evaluated three deep learning approaches for the segmenting individual LN levels I–V, which were manually contoured on CT scans from 70 HNC patients. The networks were trained and evaluated using five-fold cross-validation and ensemble learning for 60 patients with (1) 3D patch-based UNets, (2) multi-view (MV) voxel classification networks and (3) sequential UNet+MV. The performances were evaluated using Dice similarity coefficients (DSC) for automated and manual segmentations for individual levels, and the planning target volumes were extrapolated from the combined levels I–V and II–IV, both for the cross-validation and for an independent test set of 10 patients. The median DSC were 0.80, 0.66 and 0.82 for UNet, MV and UNet+MV, respectively. Overall, UNet+MV significantly (<i>p</i> < 0.0001) outperformed other arrangements and yielded DSC = 0.87, 0.85, 0.86, 0.82, 0.77, 0.77 for the combined and individual level I–V structures, respectively. Both PTVs were also significantly (<i>p</i> < 0.0001) more accurate with UNet+MV, with DSC = 0.91 and 0.90, respectively. The accurate segmentation of individual LN levels I–V can be achieved using an ensemble of UNets. UNet+MV can further refine this result. |
first_indexed | 2024-03-09T18:26:32Z |
format | Article |
id | doaj.art-8c3a154519354a02809522bb9b7238e9 |
institution | Directory Open Access Journal |
issn | 2072-6694 |
language | English |
last_indexed | 2024-03-09T18:26:32Z |
publishDate | 2022-11-01 |
publisher | MDPI AG |
record_format | Article |
series | Cancers |
spelling | doaj.art-8c3a154519354a02809522bb9b7238e92023-11-24T07:52:12ZengMDPI AGCancers2072-66942022-11-011422550110.3390/cancers14225501Deep Learning for Automated Elective Lymph Node Level Segmentation for Head and Neck Cancer RadiotherapyVictor I. J. Strijbis0Max Dahele1Oliver J. Gurney-Champion2Gerrit J. Blom3Marije R. Vergeer4Berend J. Slotman5Wilko F. A. R. Verbakel6Department of Radiation Oncology, Amsterdam UMC Location Vrije Universiteit Amsterdam, De Boelelaan 1117, 1081 HV Amsterdam, The NetherlandsDepartment of Radiation Oncology, Amsterdam UMC Location Vrije Universiteit Amsterdam, De Boelelaan 1117, 1081 HV Amsterdam, The NetherlandsDepartment of Radiology and Nuclear Medicine, Amsterdam UMC Location University of Amsterdam, Meibergdreef 9, 1105 AZ Amsterdam, The NetherlandsDepartment of Radiation Oncology, Amsterdam UMC Location Vrije Universiteit Amsterdam, De Boelelaan 1117, 1081 HV Amsterdam, The NetherlandsDepartment of Radiation Oncology, Amsterdam UMC Location Vrije Universiteit Amsterdam, De Boelelaan 1117, 1081 HV Amsterdam, The NetherlandsDepartment of Radiation Oncology, Amsterdam UMC Location Vrije Universiteit Amsterdam, De Boelelaan 1117, 1081 HV Amsterdam, The NetherlandsDepartment of Radiation Oncology, Amsterdam UMC Location Vrije Universiteit Amsterdam, De Boelelaan 1117, 1081 HV Amsterdam, The NetherlandsDepending on the clinical situation, different combinations of lymph node (LN) levels define the elective LN target volume in head-and-neck cancer (HNC) radiotherapy. The accurate auto-contouring of individual LN levels could reduce the burden and variability of manual segmentation and be used regardless of the primary tumor location. We evaluated three deep learning approaches for the segmenting individual LN levels I–V, which were manually contoured on CT scans from 70 HNC patients. The networks were trained and evaluated using five-fold cross-validation and ensemble learning for 60 patients with (1) 3D patch-based UNets, (2) multi-view (MV) voxel classification networks and (3) sequential UNet+MV. The performances were evaluated using Dice similarity coefficients (DSC) for automated and manual segmentations for individual levels, and the planning target volumes were extrapolated from the combined levels I–V and II–IV, both for the cross-validation and for an independent test set of 10 patients. The median DSC were 0.80, 0.66 and 0.82 for UNet, MV and UNet+MV, respectively. Overall, UNet+MV significantly (<i>p</i> < 0.0001) outperformed other arrangements and yielded DSC = 0.87, 0.85, 0.86, 0.82, 0.77, 0.77 for the combined and individual level I–V structures, respectively. Both PTVs were also significantly (<i>p</i> < 0.0001) more accurate with UNet+MV, with DSC = 0.91 and 0.90, respectively. The accurate segmentation of individual LN levels I–V can be achieved using an ensemble of UNets. UNet+MV can further refine this result.https://www.mdpi.com/2072-6694/14/22/5501computed tomographydeep learninghead-and-neck cancerlymph nodesradiation oncologyauto-contouring |
spellingShingle | Victor I. J. Strijbis Max Dahele Oliver J. Gurney-Champion Gerrit J. Blom Marije R. Vergeer Berend J. Slotman Wilko F. A. R. Verbakel Deep Learning for Automated Elective Lymph Node Level Segmentation for Head and Neck Cancer Radiotherapy Cancers computed tomography deep learning head-and-neck cancer lymph nodes radiation oncology auto-contouring |
title | Deep Learning for Automated Elective Lymph Node Level Segmentation for Head and Neck Cancer Radiotherapy |
title_full | Deep Learning for Automated Elective Lymph Node Level Segmentation for Head and Neck Cancer Radiotherapy |
title_fullStr | Deep Learning for Automated Elective Lymph Node Level Segmentation for Head and Neck Cancer Radiotherapy |
title_full_unstemmed | Deep Learning for Automated Elective Lymph Node Level Segmentation for Head and Neck Cancer Radiotherapy |
title_short | Deep Learning for Automated Elective Lymph Node Level Segmentation for Head and Neck Cancer Radiotherapy |
title_sort | deep learning for automated elective lymph node level segmentation for head and neck cancer radiotherapy |
topic | computed tomography deep learning head-and-neck cancer lymph nodes radiation oncology auto-contouring |
url | https://www.mdpi.com/2072-6694/14/22/5501 |
work_keys_str_mv | AT victorijstrijbis deeplearningforautomatedelectivelymphnodelevelsegmentationforheadandneckcancerradiotherapy AT maxdahele deeplearningforautomatedelectivelymphnodelevelsegmentationforheadandneckcancerradiotherapy AT oliverjgurneychampion deeplearningforautomatedelectivelymphnodelevelsegmentationforheadandneckcancerradiotherapy AT gerritjblom deeplearningforautomatedelectivelymphnodelevelsegmentationforheadandneckcancerradiotherapy AT marijervergeer deeplearningforautomatedelectivelymphnodelevelsegmentationforheadandneckcancerradiotherapy AT berendjslotman deeplearningforautomatedelectivelymphnodelevelsegmentationforheadandneckcancerradiotherapy AT wilkofarverbakel deeplearningforautomatedelectivelymphnodelevelsegmentationforheadandneckcancerradiotherapy |