Mattsson N, et al. - In this study with 56 cognitively unimpaired controls, including 27 preclinical Alzheimer's disease (AD), 32 patients with prodromal AD, and 39 patients with AD dementia, researchers assessed the relative value of structural magnetic resonance imaging (MRI) and tau positron emission tomography (PET) as a predictor of diagnosis and cognition in AD. They used the least absolute shrinkage and selection operator with ¹⁸F-AV-1451 (tau) PET and structural MRI data (regional cortical thickness and subcortical volumes) in order to construct optimal classifiers. The diagnostic accuracy of ¹⁸F-AV-1451 in the amygdala, entorhinal cortex, parahippocampal gyrus, fusiform, and inferior parietal lobule was 93% for AD (prodromal or dementia). With 83% accuracy, the MRI classifier involved partly the same regions plus the hippocampus, however, it did not improve upon the tau classifier. Independent associations of ¹⁸F-AV-1451 retention and MRI with cognition were observed. Overall, a highly accurate AD diagnosis may be made with optimized tau PET classifiers. Partly unique information pertinent for the clinical deterioration in AD were captured by both tau PET and structural brain MRI.