The cortical signature of Alzheimer's disease: regionally specific cortical thinning relates to symptom severity in very mild to mild AD dementia and is detectable in asymptomatic amyloid-positive individuals.

This paper shows that imaging-based measurement of complex spatial patterns of brain region atrophy might capture disease-specific effects on the brain more accurately than traditional volume measures. These disease-specific spatial patterns -- a sort of disease "fingerprint" or "signature" -- represent the next step in the advancement of imaging biomarkers.

The authors performed structural MRI imaging on several cohorts of individuals and used freely available software to extract the inner and outer pial surfaces on each scan and measure cortical thickness and each position on the cortex. The per-subject cortical surfaces were aligned to each other to establish point-to-point correspondences at anatomically homologous points. In an exploratory analysis, differences in cortical thickness between patients with Alzheimer's disease (AD) and age-matched controls were mapped across the cortex, revealing a complex spatial pattern of distributed AD-associated neuronal loss that is mostly congruent with the known course of AD pathology. This AD cortical "signature" was then shown to be reproducible across different scanners and cohorts, and a secondary analysis hinted that this spatial pattern of atrophy may even be present in subjects who are not demented but nonetheless test positive for AD pathology via amyloid PET imaging with Pittsburgh compound B. The amyloid imaging results are very tentative, and the detected AD cortical signature is less than 100% congruent with what is known about AD pathology; occipital lobe, for example, appears to be damaged in the cortical signature although it is largely spared by early AD. Still, the paper presents a compelling case for the disease signature approach to MRI biomarkers as the next step toward anatomically relevant, concise, informative imaging measures.