Deep learning application for the classification of Alzheimer’s disease using 18F-flortaucipir (AV-1451) tau positron emission tomography

Subjects

A total of 271 subjects (138 CU, 75 MCI, and 63 AD) in the Alzheimer’s Disease Neuroimaging Initiative (ADNI3) for whom ¹⁸F-flortaucipir PET scans were performed at baseline were recruited. Age, sex, education, Mini-Mental State Examination (MMSE) score, ¹⁸F-flortaucipir-PET images, and diagnostic results were acquired (Fig. 3). All subjects were divided using criteria provided as clinical syndrome diagnoses within the ADNI cohort (Supplementary Table S2, S3). All subjects in the CU group had clinical dementia rating (CDR) scores of 0 or 0.5, which allowed them to be distinguished from participants with MCI and AD. The patients with MCI did not meet the dementia criteria and were evaluated based on an objective memory impairment determination. All participants with MCI had MMSE scores of 24 or higher up to 30 and CDR scores of 0.5, a CDR memory score of 0.5 or higher. In…