Evaluating the Effectiveness of Deep Learning Contouring across Multiple Radiotherapy Centres.
Auto-contouring
Deep learning contouring
Multi-centre
Organs at risk
Journal
Physics and imaging in radiation oncology
ISSN: 2405-6316
Titre abrégé: Phys Imaging Radiat Oncol
Pays: Netherlands
ID NLM: 101704276
Informations de publication
Date de publication:
Oct 2022
Oct 2022
Historique:
received:
25
01
2022
revised:
01
11
2022
accepted:
02
11
2022
entrez:
21
11
2022
pubmed:
22
11
2022
medline:
22
11
2022
Statut:
epublish
Résumé
Deep learning contouring (DLC) has the potential to decrease contouring time and variability of organ contours. This work evaluates the effectiveness of DLC for prostate and head and neck across four radiotherapy centres using a commercial system. Computed tomography scans of 123 prostate and 310 head and neck patients were evaluated. Besides one head and neck model, generic DLC models were used. Contouring time using centres' existing clinical methods and contour editing time after DLC were compared. Timing was evaluated using paired and non-paired studies. Commercial software or in-house scripts assessed dice similarity coefficient (DSC) and distance to agreement (DTA). One centre assessed head and neck inter-observer variability. The mean contouring time saved for prostate structures using DLC compared to the existing clinical method was 5.9 ± 3.5 min. The best agreement was shown for the femoral heads (median DSC 0.92 ± 0.03, median DTA 1.5 ± 0.3 mm) and the worst for the rectum (median DSC 0.68 ± 0.04, median DTA 4.6 ± 0.6 mm). The mean contouring time saved for head and neck structures using DLC was 16.2 ± 8.6 min. For one centre there was no DLC time-saving compared to an atlas-based method. DLC contours reduced inter-observer variability compared to manual contours for the brainstem, left parotid gland and left submandibular gland. Generic prostate and head and neck DLC models can provide time-savings which can be assessed with paired or non-paired studies to integrate with clinical workload. Reducing inter-observer variability potential has been shown.
Sections du résumé
Background and purpose
UNASSIGNED
Deep learning contouring (DLC) has the potential to decrease contouring time and variability of organ contours. This work evaluates the effectiveness of DLC for prostate and head and neck across four radiotherapy centres using a commercial system.
Materials and methods
UNASSIGNED
Computed tomography scans of 123 prostate and 310 head and neck patients were evaluated. Besides one head and neck model, generic DLC models were used. Contouring time using centres' existing clinical methods and contour editing time after DLC were compared. Timing was evaluated using paired and non-paired studies. Commercial software or in-house scripts assessed dice similarity coefficient (DSC) and distance to agreement (DTA). One centre assessed head and neck inter-observer variability.
Results
UNASSIGNED
The mean contouring time saved for prostate structures using DLC compared to the existing clinical method was 5.9 ± 3.5 min. The best agreement was shown for the femoral heads (median DSC 0.92 ± 0.03, median DTA 1.5 ± 0.3 mm) and the worst for the rectum (median DSC 0.68 ± 0.04, median DTA 4.6 ± 0.6 mm). The mean contouring time saved for head and neck structures using DLC was 16.2 ± 8.6 min. For one centre there was no DLC time-saving compared to an atlas-based method. DLC contours reduced inter-observer variability compared to manual contours for the brainstem, left parotid gland and left submandibular gland.
Conclusions
UNASSIGNED
Generic prostate and head and neck DLC models can provide time-savings which can be assessed with paired or non-paired studies to integrate with clinical workload. Reducing inter-observer variability potential has been shown.
Identifiants
pubmed: 36405563
doi: 10.1016/j.phro.2022.11.003
pii: S2405-6316(22)00093-8
pmc: PMC9668733
doi:
Types de publication
Journal Article
Langues
eng
Pagination
121-128Informations de copyright
© 2022 The Authors. Published by Elsevier B.V. on behalf of European Society of Radiotherapy & Oncology.
Déclaration de conflit d'intérêts
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Références
J Med Internet Res. 2021 Jul 12;23(7):e26151
pubmed: 34255661
Radiother Oncol. 2020 Dec;153:55-66
pubmed: 32920005
Med Phys. 2017 Feb;44(2):547-557
pubmed: 28205307
Phys Imaging Radiat Oncol. 2020 Jul 10;15:8-15
pubmed: 33458320
Br J Radiol. 2012 Aug;85(1016):e530-6
pubmed: 22815423
Radiother Oncol. 2005 Nov;77(2):182-90
pubmed: 16256231
Med Phys. 2010 Mar;37(3):970-9
pubmed: 20384233
J Med Radiat Sci. 2014 Sep;61(3):151-8
pubmed: 26229651
Br J Radiol. 2014 Nov;87(1043):20140299
pubmed: 25168198
Int J Radiat Oncol Biol Phys. 2011 Nov 15;81(4):950-7
pubmed: 20932664
Comput Biol Med. 2018 Jul 1;98:126-146
pubmed: 29787940
Radiat Oncol. 2012 Sep 18;7:160
pubmed: 22989046
Radiother Oncol. 2014 Sep;112(3):321-5
pubmed: 25216572
Strahlenther Onkol. 2012 Sep;188(9):807-15
pubmed: 22669393
Radiat Oncol. 2012 Mar 13;7:32
pubmed: 22414264
Int J Radiat Oncol Biol Phys. 2019 Jul 1;104(3):677-684
pubmed: 30836167
Phys Imaging Radiat Oncol. 2019 Dec 17;13:1-6
pubmed: 33458300
Pract Radiat Oncol. 2021 Jan-Feb;11(1):e80-e89
pubmed: 32599279
Diagnostics (Basel). 2020 Nov 17;10(11):
pubmed: 33212793
JAMA Netw Open. 2020 Nov 2;3(11):e2027426
pubmed: 33252691
Radiother Oncol. 2018 Feb;126(2):312-317
pubmed: 29208513
Phys Imaging Radiat Oncol. 2020 Oct 14;16:54-60
pubmed: 33458344
Radiother Oncol. 2020 Jan;142:115-123
pubmed: 31653573
Radiother Oncol. 2015 Oct;117(1):83-90
pubmed: 26277855
Radiat Oncol. 2021 Jun 8;16(1):101
pubmed: 34103062