Ensemble CNN and Uncertainty Modeling to Improve Automatic Identification/Segmentation of Multiple Sclerosis Lesions in Magnetic Resonance Imaging

TL;DR

This paper introduces an ensemble CNN framework with uncertainty modeling for MS lesion segmentation in MRI, achieving performance comparable to human experts by emulating diagnostic uncertainty handling.

Contribution

It presents a novel ensemble CNN approach with uncertainty modeling and multi-directional training to improve MS lesion segmentation accuracy.

Findings

Automated method matches human expert performance.

Ensemble classifier effectively merges multi-directional CNN outputs.

Framework validated on public MSSEG dataset with promising results.

Abstract

To date, several automated strategies for identification/segmentation of Multiple Sclerosis (MS) lesions with the use of Magnetic Resonance Imaging (MRI) have been presented, but they are outperformed by human experts, from whom they act very differently. This is mainly due to: the ambiguity originated by MRI instabilities; peculiar variability of MS; non specificity of MRI regarding MS. Physicians partially manage the uncertainty generated by ambiguity relying on radiological/clinical/anatomical background and experience. To emulate human diagnosis, we present an automated framework for identification/segmentation of MS lesions from MRI based on three pivotal concepts: 1. the modelling of uncertainty; 2. the proposal of two, separately trained, CNN, one optimized for lesions and the other for lesions with respect to the environment surrounding them, respectively repeated for axial, coronal and sagittal directions; 3. the definition of an ensemble classifier to merge the information collected by different CNN. The proposed framework is trained, validated and tested on the 2016 MSSEG benchmark public data set from a single imaging modality, the FLuid-Attenuated Inversion Recovery (FLAIR). The comparison with the ground-truth and with each of 7 human raters, proves that there is no significant difference between the automated and the human raters.