Machine learning for prediction of pathology based on neuropsychological performance and lifestyle data Lead Investigator: Thanaphong Phongpreecha Institution : Stanford University E-Mail : tpjoe@stanford.edu Proposal ID : 1671 Proposal Description: The hypothesis we want to test is if there is a pattern in neuropsychological performance and lifestyle data that could be leveraged by machine learning models to predict pathological features, particularly neuropathology such as Braak scores or CERAD. If successful, this research would be impactful because it would establish a relationship, and hence prediction, between features that are measurable when the patients are still alive to the true neuropathology of Alzheimer's disease that could only be obtained from autopsy. As a secondary goal, we would also aim to establish a relationship between neuropsychological and lifestyle data to the biomarker values (CSF and MRI).