Search

Artificial intelligence has recently gained popularity across different medical fields to aid in the detection of diseases based on pathology samples or medical imaging findings. Brain magnetic resonance imaging (MRI) is a key assessment tool for patients with temporal lobe epilepsy (TLE). The role of machine learning and artificial intelligence to increase detection of brain abnormalities in TLE remains inconclusive. We used support vector machine (SV) and deep learning (DL) models based on region of interest (ROI-based) structural (n = 336) and diffusion (n = 863) brain MRI data from patients with TLE with (“lesional”) and without (“non-lesional”) radiographic features suggestive of underlying hippocampal sclerosis from the multinational (multi-center) ENIGMA-Epilepsy consortium. Our data showed that models to identify TLE performed better or similar (68–75%) compared to models to lateralize the side of TLE (56–73%, except structural-based) based on diffusion data with the opposite pattern seen for structural data (67–75% to diagnose vs. 83% to lateralize). In other aspects, structural and diffusion-based models showed similar classification accuracies. Our classification models for patients with hippocampal sclerosis were more accurate (68–76%) than models that stratified non-lesional patients (53–62%). Overall, SV and DL models performed similarly with several instances in which SV mildly outperformed DL. We discuss the relative performance of these models with ROI-level data and the implications for future applications of machine learning and artificial intelligence in epilepsy care.

Long term irreversible disability in multiple sclerosis (MS) is thought to be primarily driven by axonal degeneration. Axonal degeneration leads to degenerative atrophy, therefore early markers of axonal degeneration are required to predict clinical disability and treatment efficacy. Given that additional pathologies such as inflammation, demyelination and oedema are also present in MS, it is essential to develop axonal markers that are not confounded by these processes. The present study investigated a novel method for measuring axonal degeneration in MS based on high angular resolution diffusion magnetic resonance imaging. Unlike standard methods, this novel method involved advanced acquisition and modelling for improved axonal sensitivity and specificity. Recent work has developed analytical methods, two novel axonal markers, fibre density and cross-section, that can be estimated for each fibre direction in each voxel (termed a “fixel”). This technique, termed fixel-based analysis, thus simultaneously estimates axonal density and white matter atrophy from specific white matter tracts. Diffusion-weighted imaging datasets were acquired for 17 patients with a history of acute unilateral optic neuritis (35.3±10.2years, 11 females) and 14 healthy controls (32.7±4.8years, 8 females) on a 3T scanner. Fibre density values were compared to standard diffusion tensor imaging parameters (fractional anisotropy and mean diffusivity) in lesions and normal appearing white matter. Group comparisons were performed for each fixel to assess putative differences in fibre density and fibre cross-section. Fibre density was observed to have a comparable sensitivity to fractional anisotropy for detecting white matter pathology in MS, but was not affected by crossing axonal fibres. Whole brain fixel-based analysis revealed significant reductions in fibre density and fibre cross-section in the inferior fronto-occipital fasciculus (including the optic radiations) of patients compared to controls. We interpret this result to indicate that this fixel-based approach is able to detect early loss of fibre density and cross-section in the optic radiations in MS patients with a history of optic neuritis. Fibre-specific markers of axonal degeneration should be investigated further for use in early stage therapeutic trials, or to monitor axonal injury in early stage MS. Keywords: Multiple sclerosis, MRI, Diffusion, Optic neuritis, Axonal degeneration, Apparent fibre density