Abstract
AbstractArtificial intelligence (AI)-based tools are widely employed, but their use for diagnosis and prognosis of neurological disorders is still evolving. We capitalise on a large-scale, cross-sectional structural MRI dataset of 814 people with epilepsy. We use a recently developed machine-learning algorithm, Subtype and Stage Inference (SuStaIn), to develop a novel data-driven disease taxonomy based on distinct patterns of spatiotemporal progression of brain atrophy. We identify two subtypes common to focal and idiopathic generalised epilepsies, characterised by neocortical-driven or basal ganglia-driven progression, and a third subtype, only detected in focal epilepsies, characterised by hippocampus-driven progression. We corroborate external validity via an independent cohort of 254 people and decode associations between progression subtypes and clinical measures of epilepsy severity. Our findings suggest fundamental processes underlying the progression of epilepsy-related brain atrophy. We deliver a novel MRI- and AI-guided epilepsy taxonomy, which could be used for individualised prognostics and targeted therapeutics.
Publisher
Cold Spring Harbor Laboratory