The use of the NACC database to evaluate a clinical decisions support tool in dementia Lead Investigator: Antti Tolonen Institution : Combinostics E-Mail : antti.tolonen@combinostics.com Proposal ID : 480 Proposal Description: ** RSG Note: this was previously under Lennart Thurfjell who is still indirectly involved in this project, but the main analysis is now under Mr. Tolonen** Computerized tools in the diagnostics of neurodegenerative diseases are still very rare. Integrating information from multiple sources, such as imaging biomarkers and results from clinical and neuropsychological tests, is challenging. We have developed a tool that integrates this type of heterogeneous data and provides the physician with a holistic view of the patient???s disease state. We compute a Disease State Fingerprint (DSF), by comparing the patient???s test result with distributions based on previously diagnosed patients and by combining the results of all tests in a hierarchical fashion. The purpose of the proposed project is to use the NACC database to build DSF distributions based on normal, AD, FTD, LBD and VD cases and to evaluate the usefulness of the DSF method for disease classification. We will use cross-validation to compute the accuracy of the DSF classification as compared to the clinical diagnosis from follow up visits.