Distance-based detection of out-of-distribution silent failures for Covid-19 lung lesion segmentation

TL;DR

This paper introduces a simple Mahalanobis distance-based method for detecting out-of-distribution failures in medical image segmentation, improving trustworthiness in clinical settings.

Contribution

A lightweight OOD detection approach that integrates into existing segmentation models using Mahalanobis distance, applicable across various medical imaging modalities.

Findings

Effectively detects far- and near-OOD samples in chest CT scans.

Validates across multiple distribution shifts and MRI applications.

Enhances clinical reliability of segmentation models.

Abstract

Automatic segmentation of ground glass opacities and consolidations in chest computer tomography (CT) scans can potentially ease the burden of radiologists during times of high resource utilisation. However, deep learning models are not trusted in the clinical routine due to failing silently on out-of-distribution (OOD) data. We propose a lightweight OOD detection method that leverages the Mahalanobis distance in the feature space and seamlessly integrates into state-of-the-art segmentation pipelines. The simple approach can even augment pre-trained models with clinically relevant uncertainty quantification. We validate our method across four chest CT distribution shifts and two magnetic resonance imaging applications, namely segmentation of the hippocampus and the prostate. Our results show that the proposed method effectively detects far- and near-OOD samples across all explored scenarios.