Estimating Uncertainty in Neural Networks for Cardiac MRI Segmentation: A Benchmark Study

by   Matthew Ng, et al.

Convolutional neural networks (CNNs) have demonstrated promise in automated cardiac magnetic resonance imaging segmentation. However, when using CNNs in a large real world dataset, it is important to quantify segmentation uncertainty in order to know which segmentations could be problematic. In this work, we performed a systematic study of Bayesian and non-Bayesian methods for estimating uncertainty in segmentation neural networks. We evaluated Bayes by Backprop (BBB), Monte Carlo (MC) Dropout, and Deep Ensembles in terms of segmentation accuracy, probability calibration, uncertainty on out-of-distribution images, and segmentation quality control. We tested these algorithms on datasets with various distortions and observed that Deep Ensembles outperformed the other methods except for images with heavy noise distortions. For segmentation quality control, we showed that segmentation uncertainty is correlated with segmentation accuracy. With the incorporation of uncertainty estimates, we were able to reduce the percentage of poor segmentation to 5 manual review, substantially lower than random review of the results without using neural network uncertainty.


page 6

page 7

page 8

page 9

page 15

page 16


Bayesian QuickNAT: Model Uncertainty in Deep Whole-Brain Segmentation for Structure-wise Quality Control

We introduce Bayesian QuickNAT for the automated quality control of whol...

Temporal Uncertainty Localization to Enable Human-in-the-loop Analysis of Dynamic Contrast-enhanced Cardiac MRI Datasets

Dynamic contrast-enhanced (DCE) cardiac magnetic resonance imaging (CMRI...

Ensemble of Pre-Trained Neural Networks for Segmentation and Quality Detection of Transmission Electron Microscopy Images

Automated analysis of electron microscopy datasets poses multiple challe...

Bayesian Neural Networks for Uncertainty Estimation of Imaging Biomarkers

Image segmentation enables to extract quantitative measures from scans t...

Uncertainty Estimation in Deep 2D Echocardiography Segmentation

2D echocardiography is the most common imaging modality for cardiovascul...

Please sign up or login with your details

Forgot password? Click here to reset