Primary Diagnosis Prediction from Chief Complaints

摘要

Telemedicine is one of the primary health policies in Thailand to improve life quality of people by accessing medical service from anywhere anytime. Many rural areas still have difficulties to get medical services such as primary diagnosis. We propose a web-based application for chief complaint analysis that is publicly available. Our system takes input of chief complaint text then makes predictions of primary diagnosis using machine learning algorithm. Our experiment shows that the model has promising results at 0.77 of accuracy. Our system could be used as a patient screening tool that makes impact to operational excellence for Thailand healthcare.