A Heteroscedastic Uncertainty Model for Decoupling Sources of MRI Image Quality

Richard Shaw, Carole H. Sudre, S. bastien Ourselin, M. Jorge Cardoso

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

3 Citations (Scopus)

Abstract

Quality control (QC) of medical images is essential to ensure that downstream analyses such as segmentation can be performed successfully. Currently, QC is predominantly performed visually at significant time and operator cost. We aim to automate the process by formulating a probabilistic network that estimates uncertainty through a heteroscedastic noise model, hence providing a proxy measure of task-specific image quality that is learnt directly from the data. By augmenting the training data with different types of simulated k-space artefacts, we propose a novel cascading CNN architecture based on a student-teacher framework to decouple sources of uncertainty related to different k-space augmentations in an entirely self-supervised manner. This enables us to predict separate uncertainty quantities for the different types of data degradation. While the uncertainty measures reflect the presence and severity of image artefacts, the network also provides the segmentation predictions given the quality of the data. We show models trained with simulated artefacts provide informative measures of uncertainty on real-world images and we validate our uncertainty predictions on problematic images identified by human-raters.
Original languageEnglish
Title of host publicationProceedings of the 3rd Conference on Medical Imaging with Deep Learning, MIDL 2020
EditorsTal Arbel, Ismail Ben Ayed, Marleen de Bruijne, Maxime Descoteaux, Herve Lombaert, Christopher Pal
PublisherML Research Press
Pages733-742
Volume121
Publication statusPublished - 2020
Event3rd Conference on Medical Imaging with Deep Learning, MIDL 2020 - Virtual, Online, Canada
Duration: 6 Jul 20208 Jul 2020

Publication series

NameProceedings of Machine Learning Research

Conference

Conference3rd Conference on Medical Imaging with Deep Learning, MIDL 2020
Country/TerritoryCanada
CityVirtual, Online
Period6/07/20208/07/2020

Cite this