Providemus alz: Ubiquitous Preclinical Alzheimer’s Disease Screening
Short abstract: Brain health changes throughout adulthood, yet conventional assessments capture only isolated moments. This thesis investigates whether everyday data from consumer wearables and smartphones can complement these assessments by characterizing cognitive performance and affective functioning over time. Through the two-year Providemus alz study, adults aged 45 and older contributed continuous activity, sleep, and heart-rate data alongside environmental information and repeated cognitive and affective assessments. Across a systematic review and empirical studies, findings demonstrate the feasibility of sustained monitoring and show that passive signals contain meaningful predictive information about multiple brain-health outcomes. However, their predictive value varies across individuals, outcomes, and time windows. Models also supported forecasting of cognitive and affective functioning up to six months ahead, although performance differed between participants. These findings establish important methodological foundations for individualized frequent, ubiquitous, and fully passive brain-health monitoring, while underscoring that passive sensing complements rather than replaces clinical assessments.
Igor Matias
Co-supervisors: Prof. Dr. Katarzyna Wac, Prof. Dr. Matthias Kliegel
In Archive Ouverte of UNIGE.
https://doi.org/10.13097/archive-ouverte/unige:197282

