PROJECT SUMMARY Alzheimer's disease and related dementias (AD/ADRD) rank prominently among age-associated neurodegenerative disorders, with Parkinson’s disease trailing as the second most common. Contemporary studies on AD and Mild Cognitive Impairment (MCI) have unveiled cognitive decline indicators mirrored in nuances within gait and hand movements. These indications often emerge long before AD or MCI diagnoses are confirmed. Consistently, research has correlated slowed gait with cognitive deterioration, elevated brain amyloid levels, and an augmented AD risk. However, the interplay between systems influencing both cognition and movement has largely been explored in separate studies. The depth of understanding around the cognitive efforts required for gait initiation or motor planning remains scant. Though dementia screening traditionally hinges on comprehensive neuropsychological evaluations and neuroimaging biomarker studies, gait and movement assessment offers a more direct approach. Crucially, motor features remain unaffected by language, educational background, or cognitive capabilities, positioning them as unbiased and consistent evaluative instruments. For example, the Short Physical Performance Battery (SPPB) tests can be easily conducted at home and monitored, even using smartphone applications. In this project, we chart a course to harness cutting-edge AI and computer vision (CV) tools to aid in distinguishing diverse dementia and MCI subtypes, such as AD-related MCI versus MCI with PD pathologies. Our approach integrates multi-modal brain imaging with gait and movement data from videos, aiming to extract nuanced markers for phenotyping (pre)clinical AD. These markers can then serve as precise digital trackers for both the progression and distinction of age-related degenerative disorders. Ultimately, we aspire to enhance the understanding of the intricate links between gait, movement, and cognition in aging and AD/ADRD scenarios.