Laplacian Segmentation Networks Improve Epistemic Uncertainty Quantification

Image segmentation relies heavily on neural networks which are known to be overconfident, especially when making predictions on out-of-distribution (OOD) images. This is a common scenario in the medical domain due to variations in equipment, acquisition sites, or image corruptions. This work address...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2024-07
Main Authors: Zepf, Kilian, Wanna, Selma, Miani, Marco, Moore, Juston, Frellsen, Jes, Hauberg, Søren, Warburg, Frederik, Feragen, Aasa
Format: Article
Language:English
Subjects:
Balances (scales)
Image segmentation
Medical imaging
Neural networks
Uncertainty
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Image segmentation relies heavily on neural networks which are known to be overconfident, especially when making predictions on out-of-distribution (OOD) images. This is a common scenario in the medical domain due to variations in equipment, acquisition sites, or image corruptions. This work addresses the challenge of OOD detection by proposing Laplacian Segmentation Networks (LSN): methods which jointly model epistemic (model) and aleatoric (data) uncertainty for OOD detection. In doing so, we propose the first Laplace approximation of the weight posterior that scales to large neural networks with skip connections that have high-dimensional outputs. We demonstrate on three datasets that the LSN-modeled parameter distributions, in combination with suitable uncertainty measures, gives superior OOD detection.
ISSN:2331-8422