Functional neuroimaging has significantly informed us about healthy brains and their commonalities at group level. Yet, human cognitive functions, and also the ways in which these functions decay in disease, are highly individual. Yet, biomarkers for estimating the individual course of many neurological diseases are sorely lacking. Here we combine functional neuroimaging to novel computational analysis tools, for improving clinical diagnostics and prognostic estimations in stroke, Parkinson's disease, and mild traumatic brain injury. We will measure magnetoencephalographic (MEG) responses longitudinally in patients and healthy controls, and utilize our recent advances in computational approaches. We aim at i) providing individual and easily applicable neuroimaging measures to be used as functional biomarkers for estimating the disease trajectories in these patients, and ii) obtaining information of the usability of different neuroimaging measures as clinical tools.