Ontology-supported knowledge management with case-based reason-ing for intelligent health projection

Abstract:
For the past few years, health-related issues are getting more and more attention. With regular health screening, disease detection and subsequent medical attention can proceed with far lesser chance of missing the critical period. The projection of health related issues from health screening has been studied by using statistical analysis methods. However, the results generally fail to alert the subjects regarding the potential risks involved due to lack of certainty. The potential health issues derived from statistical analysis of health screening data and medical general guidelines can only be presented to the subjects with probability and lack of supportive hard evidence. The efficacy of health screening consequently fails to be realized. In order to confront the dilemma, we have developed an Intelligent Health Projection System (IHPS) for providing an evidential health status projection and a more motivated health plan to the subjects. This study focuses on modelling Type 2 Diabetes Mellitus (T2DM) and associated factors as an example. Other cases can be analogously implemented. The IHPS adaptively provides the projection of potential risk of getting T2DM by exploring the similarity between the subject’s health screening data and previous T2DM patients’ cases. The main building blocks, the cases that serve as a knowledge base for IHPS are modelled using ontology technology. As the main functionality of IHPS, the T2DM projection uses the traces left by previous T2DM patients and works on top of case-based reasoning (CBR) mechanism. The proposed IHPS aims to promote better self-health management by enhancing a subject’s comprehension on risks revealed in health screening result. The cases retrieved can not only being used for risk projection of a subject but also serving as evidences for physicians to provide more accurate and convincing health advice.