Abstract

Motivation: In multiple sclerosis, slowly expanding lesions have been suggested as a hallmark of a steadily worsening disease course. However, identifying these lesions is challenging, as their growth rates are at the detection limit of today's processing algorithms or MRI data must be available over a long period of time. Goal(s): To identify and characterise slowly expanding lesions in cross-sectional data. Approach: We compared changes in quantitative T1, T2 and T2/T1-ratio inside lesions and in perilesional tissue for enlarging/stable/shrinking/new lesion phenotypes. Results: Z-scores of multiparametric quantitative maps carry discriminative information to classify lesion evolution from single time point data. Impact: Our findings suggest that quantitative multiparametric analyses allow a better in vivo characterisation of microstructural tissue pathology in multiple sclerosis; this furthers the understanding of different lesion evolutions and might enable to already distinguish them from cross-sectional data.