Please use this identifier to cite or link to this item: http://dx.doi.org/10.25673/79508
Title: Domain adaptation for segmentation of critical structures for prostate cancer therapy
Author(s): Meyer, Anneke
Mehrtash, Alireza
Rak, MarkoLook up in the Integrated Authority File of the German National Library
Bashkanov, Oleksii
Langbein, Bjoern
Ziaei, Alireza
Kibel, Adam S.
Tempany, Clare M.
Hansen, ChristianLook up in the Integrated Authority File of the German National Library
Tokuda, Junichi
Issue Date: 2021
Type: Article
Language: English
URN: urn:nbn:de:gbv:ma9:1-1981185920-814622
Subjects: Prostate cancer
Therapy
Domain adaptation
Segmentation
Abstract: Preoperative assessment of the proximity of critical structures to the tumors is crucial in avoiding unnecessary damage during prostate cancer treatment. A patient-specific 3D anatomical model of those structures, namely the neurovascular bundles (NVB) and the external urethral sphincters (EUS), can enable physicians to perform such assessments intuitively. As a crucial step to generate a patient-specific anatomical model from preoperative MRI in a clinical routine, we propose a multi-class automatic segmentation based on an anisotropic convolutional network. Our specific challenge is to train the network model on a unique source dataset only available at a single clinical site and deploy it to another target site without sharing the original images or labels. As network models trained on data from a single source suffer from quality loss due to the domain shift, we propose a semi-supervised domain adaptation (DA) method to refine the model’s performance in the target domain. Our DA method combines transfer learning and uncertainty guided self-learning based on deep ensembles. Experiments on the segmentation of the prostate, NVB, and EUS, show significant performance gain with the combination of those techniques compared to pure TL and the combination of TL with simple self-learning ( p < 0.005 for all structures using a Wilcoxon’s signed-rank test). Results on a different task and data (Pancreas CT segmentation) demonstrate our method’s generic application capabilities. Our method has the advantage that it does not require any further data from the source domain, unlike the majority of recent domain adaptation strategies. This makes our method suitable for clinical applications, where the sharing of patient data is restricted.
URI: https://opendata.uni-halle.de//handle/1981185920/81462
http://dx.doi.org/10.25673/79508
Open Access: Open access publication
License: (CC BY-SA 4.0) Creative Commons Attribution ShareAlike 4.0(CC BY-SA 4.0) Creative Commons Attribution ShareAlike 4.0
Sponsor/Funder: OVGU-Publikationsfonds 2021
Journal Title: Scientific reports
Publisher: Macmillan Publishers Limited, part of Springer Nature
Publisher Place: [London]
Volume: 11
Issue: 2021
Original Publication: 10.1038/s41598-021-90294-4
Page Start: 1
Page End: 14
Appears in Collections:Fakultät für Informatik (OA)

Files in This Item:
File Description SizeFormat 
Meyer et al._Domain adaptation_2021.pdfZweitveröffentlichung2.48 MBAdobe PDFThumbnail
View/Open