Bitte benutzen Sie diese Kennung, um auf die Ressource zu verweisen: https://doi.org/10.21256/zhaw-25502
Publikationstyp: Beitrag in wissenschaftlicher Zeitschrift
Art der Begutachtung: Peer review (Publikation)
Titel: Unsupervised domain adaptation for vertebrae detection and identification in 3D CT volumes using a domain sanity loss
Autor/-in: Sager, Pascal
Salzmann, Sebastian
Burn, Felice
Stadelmann, Thilo
et. al: No
DOI: 10.3390/jimaging8080222
10.21256/zhaw-25502
Erschienen in: Journal of Imaging
Band(Heft): 8
Heft: 8
Seite(n): 222
Erscheinungsdatum: 19-Aug-2022
Verlag / Hrsg. Institution: MDPI
ISSN: 2313-433X
Sprache: Englisch
Schlagwörter: Unsupervised domain adaptation; Semi-supervised learning; Vertebrae detection; Vertebrae identification; Transfer learning; Semantic segmentation; Data centrism; Deep learning
Fachgebiet (DDC): 006: Spezielle Computerverfahren
616: Innere Medizin und Krankheiten
Zusammenfassung: A variety of medical computer vision applications analyze 2D slices of computed tomography (CT) scans, whereas axial slices from the body trunk region are usually identified based on their relative position to the spine. A limitation of such systems is that either the correct slices must be extracted manually or labels of the vertebrae are required for each CT scan to develop an automated extraction system. In this paper, we propose an unsupervised domain adaptation (UDA) approach for vertebrae detection and identification based on a novel Domain Sanity Loss (DSL) function. With UDA the model’s knowledge learned on a publicly available (source) data set can be transferred to the target domain without using target labels, where the target domain is defined by the specific setup (CT modality, study protocols, applied pre- and processing) at the point of use (e.g., a specific clinic with its specific CT study protocols). With our approach, a model is trained on the source and target data set in parallel. The model optimizes a supervised loss for labeled samples from the source domain and the DSL loss function based on domain-specific “sanity checks” for samples from the unlabeled target domain. Without using labels from the target domain, we are able to identify vertebra centroids with an accuracy of 72.8%. By adding only ten target labels during training the accuracy increases to 89.2%, which is on par with the current state-of-the-art for full supervised learning, while using about 20 times less labels. Thus, our model can be used to extract 2D slices from 3D CT scans on arbitrary data sets fully automatically without requiring an extensive labeling effort, contributing to the clinical adoption of medical imaging by hospitals.
URI: https://digitalcollection.zhaw.ch/handle/11475/25502
Volltext Version: Publizierte Version
Lizenz (gemäss Verlagsvertrag): CC BY 4.0: Namensnennung 4.0 International
Departement: School of Engineering
Organisationseinheit: Centre for Artificial Intelligence (CAI)
Enthalten in den Sammlungen:Publikationen School of Engineering

Dateien zu dieser Ressource:
Datei Beschreibung GrößeFormat 
2022_Sager-etal_Unsupervised-domain-adaptation-vertebrae-detection-3D-CT.pdf32.91 MBAdobe PDFMiniaturbild
Öffnen/Anzeigen
Zur Langanzeige


Alle Ressourcen in diesem Repository sind urheberrechtlich geschützt, soweit nicht anderweitig angezeigt.