Project
We aim to employ recent advances in computer vision and artificial intelligence to assist in differentiating dementia phenotypes with a solid understanding of the neural underpinnings of gait and mobility disruptions related to dementia. Our analysis will identify personalized digital biomarkers by analyzing 3D movement patterns in videos (e.g., taken by iPads) to derive a set of interpretable features. We assess the accuracy and reliability of these gait-derived markers to predict gait and mobility impairment and further link them to underlying neural mechanisms and clinical characteristics. By utilizing these markers, we can track the progression of the disease, potentially reducing the need for costly neuropsychological or neuroimaging assessments. Data for this study comes from the participants at the Stanford AD Research Center (ADRC) and will include structural and functional MRI scans, diffusion tensor imaging, and assessments of motor capabilities using the Short Performance Physical Battery (SPPB), Movement Disorders Society-Unified Parkinson's Disease Rating Scale (MDS-UPDRS), Luria’s three-step tests. Controlled comparisons will be conducted between healthy controls, Mild Cognitive Impairment due to Alzheimer's Disease (MCIAD), and AD dementia, as well as between MCIAD and MCI with Parkinson’s Disease (MCIPD) pathologies to specify phenotypes. Selected brain connectome features (structural, functional, or coupling) will strengthen the specificity of video-recorded gait and balance features for differentiating MCIAD from MCIPD since the trajectory of an individual’s gait variabilities, emerged because of cognitive-cortical dysfunction, differs between AD and PD etiologies and their subtypes. Our analysis will have enduring impacts by uncovering neural pathways related to gait and mobility, thereby deepening our understanding of the fundamental mechanisms involved. This approach not only improves early detection and monitoring but also enhances our ability to differentiate between neural mechanisms of AD pathologies and other neurodegenerative diseases, leading to more targeted and effective interventions.
