StRegA: Unsupervised anomaly detection in brain MRIs using a compact context-encoding variational autoencoder

Expert interpretation of anatomical images of the human brain is the central part of neuroradiology. Several machine learning-based techniques have been proposed to assist in the analysis process. However, the ML models typically need to be trained to perform a specific task, e.g., brain tumour segm...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Computers in biology and medicine Jg. 149; S. 106093
Hauptverfasser: Chatterjee, Soumick, Sciarra, Alessandro, Dünnwald, Max, Tummala, Pavan, Agrawal, Shubham Kumar, Jauhari, Aishwarya, Kalra, Aman, Oeltze-Jafra, Steffen, Speck, Oliver, Nürnberger, Andreas
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Oxford Elsevier Ltd 01.10.2022
Elsevier Limited
Schlagworte:
Abnormalities
Annotations
Anomalies
Anomaly detection
Brain
Brain cancer
Brain tumors
Brain tumour segmentation
Context
Datasets
Image segmentation
Internal Medicine
Machine learning
Medical imaging
MRI
Neuroimaging
Other
Training
Tumors
Unsupervised learning
Brain tumour segmentation
MRI
Anomaly detection
Unsupervised learning
ISSN:0010-4825, 1879-0534, 1879-0534
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Expert interpretation of anatomical images of the human brain is the central part of neuroradiology. Several machine learning-based techniques have been proposed to assist in the analysis process. However, the ML models typically need to be trained to perform a specific task, e.g., brain tumour segmentation or classification. Not only do the corresponding training data require laborious manual annotations, but a wide variety of abnormalities can be present in a human brain MRI — even more than one simultaneously, which renders a representation of all possible anomalies very challenging. Hence, a possible solution is an unsupervised anomaly detection (UAD) system that can learn a data distribution from an unlabelled dataset of healthy subjects and then be applied to detect out-of-distribution samples. Such a technique can then be used to detect anomalies — lesions or abnormalities, for example, brain tumours, without explicitly training the model for that specific pathology. Several Variational Autoencoder (VAE) based techniques have been proposed in the past for this task. Even though they perform very well on controlled artificially simulated anomalies, many of them perform poorly while detecting anomalies in clinical data. This research proposes a compact version of the “context-encoding” VAE (ceVAE) model, combined with pre and post-processing steps, creating a UAD pipeline (StRegA), which is more robust on clinical data and shows its applicability in detecting anomalies such as tumours in brain MRIs. The proposed pipeline achieved a Dice score of 0.642 ± 0.101 while detecting tumours in T2w images of the BraTS dataset and 0.859 ± 0.112 while detecting artificially induced anomalies, while the best performing baseline achieved 0.522 ± 0.135 and 0.783 ± 0.111, respectively. •Proposes unsupervised anomaly detection pipeline StRegA and compact ceVAE model.•The model is combined the proposed pre- and post-processing steps to form StRegA.•Trained on anomaly-free brain MRI datasets and evaluated for the task of brain tumour detection.•Proposed pipeline achieved a Dice score of 0.642 ± 0.101 while detecting tumours in T2w images of the BraTS dataset.•Achieved a Dice score of 0.859 ± 0.112 while detecting artificially induced anomalies.
Bibliographie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:0010-4825
1879-0534
1879-0534
DOI:10.1016/j.compbiomed.2022.106093