Abstract:Alzheimer's disease (AD) is a chronic neurodegenerative condition responsible for most cases of dementia and considered as one of the greatest challenges for neuroscience in this century. Early Ad signs are usually mistaken for normal age-related cognitive dysfunctions, thus patients usually start their treatment in advanced AD stages, when its benefits are severely limited. AD has no known cure, as such, hope lies on early diagnosis which usually depends on neuroimaging techniques such as Positron Emission Tomography (PET). PET data is then analyzed with Statistical Parametric Mapping (SPM) software, which uses mass univariate statistical analysis, inevitably incurring in errors derived from this multiple testing approach. Recently, Wang et al. (2020) formulated an alternative: applying functional data analysis (FDA), a relatively new branch of statistics, to calculate mean function and simultaneous confidence corridors (SCCs) for the difference between two groups' PET values. Here we test this approach with a practical application for AD diagnosis, estimating mean functions and SCCs for the difference between AD and control group's PET activity and locating regions where this difference galls outside estimated SCCs, indicating differences in brain activity attributable to AD-derived neural loss. Our results are consistent with previous literature on AD pathology and suggest that this FDA approach is more resilient to reductions in sample size and less dependent on ad hoc selection of an {\alpha} level than its counterpart, suggesting that this novel technique is a promising venue for research in the field of medical imaging.

Simultaneous Confidence Corridors for neuroimaging data analysis: applications to Alzheimer's Disease diagnosis