Exploring Uncertainty Measures in Deep Networks for Multiple Sclerosis Lesion Detection and Segmentation

TL;DR

This paper investigates multiple uncertainty estimation methods using Monte Carlo dropout in deep learning models for detecting and segmenting multiple sclerosis lesions in MRI scans, aiming to improve clinical decision-making.

Contribution

It introduces the first application of multiple MC dropout-based uncertainty measures in MS lesion segmentation, enhancing model reliability and clinical utility.

Findings

Uncertainty measures improve lesion detection and segmentation performance.

Operating points based on uncertainty outperform those based solely on sigmoid outputs.

The approach provides voxel- and lesion-level uncertainty estimates for clinical assessment.

Abstract

Deep learning (DL) networks have recently been shown to outperform other segmentation methods on various public, medical-image challenge datasets [3,11,16], especially for large pathologies. However, in the context of diseases such as Multiple Sclerosis (MS), monitoring all the focal lesions visible on MRI sequences, even very small ones, is essential for disease staging, prognosis, and evaluating treatment efficacy. Moreover, producing deterministic outputs hinders DL adoption into clinical routines. Uncertainty estimates for the predictions would permit subsequent revision by clinicians. We present the first exploration of multiple uncertainty estimates based on Monte Carlo (MC) dropout [4] in the context of deep networks for lesion detection and segmentation in medical images. Specifically, we develop a 3D MS lesion segmentation CNN, augmented to provide four different voxel-based uncertainty measures based on MC dropout. We train the network on a proprietary, large-scale, multi-site, multi-scanner, clinical MS dataset, and compute lesion-wise uncertainties by accumulating evidence from voxel-wise uncertainties within detected lesions. We analyze the performance of voxel-based segmentation and lesion-level detection by choosing operating points based on the uncertainty. Empirical evidence suggests that uncertainty measures consistently allow us to choose superior operating points compared only using the network's sigmoid output as a probability.