Header

UZH-Logo

Maintenance Infos

Integrating uncertainty in deep neural networks for MRI based stroke analysis


Herzog, Lisa; Murina, Elvis; Dürr, Oliver; Wegener, Susanne; Sick, Beate (2020). Integrating uncertainty in deep neural networks for MRI based stroke analysis. Medical Image Analysis, 65:101790.

Abstract

At present, the majority of the proposed Deep Learning (DL) methods provide point predictions without quantifying the model's uncertainty. However, a quantification of the reliability of automated image analysis is essential, in particular in medicine when physicians rely on the results for making critical treatment decisions. In this work, we provide an entire framework to diagnose ischemic stroke patients incorporating Bayesian uncertainty into the analysis procedure. We present a Bayesian Convolutional Neural Network (CNN) yielding a probability for a stroke lesion on 2D Magnetic Resonance (MR) images with corresponding uncertainty information about the reliability of the prediction. For patient-level diagnoses, different aggregation methods are proposed and evaluated, which combine the individual image-level predictions. Those methods take advantage of the uncertainty in the image predictions and report model uncertainty at the patient-level. In a cohort of 511 patients, our Bayesian CNN achieved an accuracy of 95.33% at the image-level representing a significant improvement of 2% over a non-Bayesian counterpart. The best patient aggregation method yielded 95.89% of accuracy. Integrating uncertainty information about image predictions in aggregation models resulted in higher uncertainty measures to false patient classifications, which enabled to filter critical patient diagnoses that are supposed to be closer examined by a medical doctor. We therefore recommend using Bayesian approaches not only for improved image-level prediction and uncertainty estimation but also for the detection of uncertain aggregations at the patient-level.

Abstract

At present, the majority of the proposed Deep Learning (DL) methods provide point predictions without quantifying the model's uncertainty. However, a quantification of the reliability of automated image analysis is essential, in particular in medicine when physicians rely on the results for making critical treatment decisions. In this work, we provide an entire framework to diagnose ischemic stroke patients incorporating Bayesian uncertainty into the analysis procedure. We present a Bayesian Convolutional Neural Network (CNN) yielding a probability for a stroke lesion on 2D Magnetic Resonance (MR) images with corresponding uncertainty information about the reliability of the prediction. For patient-level diagnoses, different aggregation methods are proposed and evaluated, which combine the individual image-level predictions. Those methods take advantage of the uncertainty in the image predictions and report model uncertainty at the patient-level. In a cohort of 511 patients, our Bayesian CNN achieved an accuracy of 95.33% at the image-level representing a significant improvement of 2% over a non-Bayesian counterpart. The best patient aggregation method yielded 95.89% of accuracy. Integrating uncertainty information about image predictions in aggregation models resulted in higher uncertainty measures to false patient classifications, which enabled to filter critical patient diagnoses that are supposed to be closer examined by a medical doctor. We therefore recommend using Bayesian approaches not only for improved image-level prediction and uncertainty estimation but also for the detection of uncertain aggregations at the patient-level.

Statistics

Citations

Dimensions.ai Metrics
17 citations in Web of Science®
16 citations in Scopus®
Google Scholar™

Altmetrics

Downloads

162 downloads since deposited on 24 Sep 2020
46 downloads since 12 months
Detailed statistics

Additional indexing

Item Type:Journal Article, refereed, original work
Communities & Collections:04 Faculty of Medicine > University Hospital Zurich > Clinic for Neurology
04 Faculty of Medicine > Neuroscience Center Zurich
Dewey Decimal Classification:570 Life sciences; biology
610 Medicine & health
Scopus Subject Areas:Health Sciences > Radiological and Ultrasound Technology
Health Sciences > Radiology, Nuclear Medicine and Imaging
Physical Sciences > Computer Vision and Pattern Recognition
Health Sciences > Health Informatics
Physical Sciences > Computer Graphics and Computer-Aided Design
Language:English
Date:October 2020
Deposited On:24 Sep 2020 05:51
Last Modified:24 Nov 2023 02:41
Publisher:Elsevier
ISSN:1361-8415
OA Status:Green
Publisher DOI:https://doi.org/10.1016/j.media.2020.101790
PubMed ID:32801096
  • Content: Accepted Version