Adaptive Medical Diagnosis Expert System (AMDES) for the Early Diagnosis of Dementia Lead Investigator: Allan Chong Institution : National University of Singapore E-Mail : kw_chong@nus.edu.sg Proposal ID : 422 Proposal Description: My project team is trying to build an expert system based on diagnostic criteria for non expert clinician / physicians on the subject of dementia. 1. What we are trying to achieve is to allow early detection of dementia by General Practitioners (GP) by providing them recommendations on the next course of actions by guiding the physician what symptoms to look out for based on rules that are derived from the DSM IV criteria. 2. What are the tests that are conducted by specialist in hospitals on their initial visits based on some data the physician already have. This is to reduce load from the hospitals so that the test results will be ready on their initial visit instead of having to order tests on initial visits at the specialist clinic. This is to allow specialist to conduct analysis of the test results on the patient's initial visit and hence saves time and improve efficiency. We have completed the first part of the project and the system is currently on pilot testing with a local hospital. However, we require some dataset for us to conduct analysis on what are the tests that are conducted on the initial visit to a specialist. We have found form E1 most useful for our study, and would like to have the UDS dataset. The data analysis will be based on machine learning techniques.