The Decision Support System Used in HEALS (Health Examination Automatic Logic System)
Chiou-Shann Fuh, PhD# Kuan-Liang Kuo, MD* #
*Family Medicine Department, Taipei City Hospital, Taipei, Taiwan
#Department of Computer Science and Information Engineering, National Taiwan University, Taipei, Taiwan
Summary: Studies have shown the importance of using clinical guidelines in medical practices. HEALS provides a guideline-based decision-support system for health examinations. A simple and clear rule syntax and an efficient reasoning algorithm are developed for HEALS in automatic interpretation of health examination results. To ensure the state-of-the-art of the knowledge base, it can be readily updated by domain experts easily. HEALS has offered great satisfaction for clinical workers and customers in several healthcare settings.
Problem addressed: There have been many efforts to build guideline-based decision-support system for medical practices. Domain divergence requires specialization in developing decision-support system. In health examination, the knowledge base dominates in covering the problem in sufficient breadth more than depth to allow sophisticated problem solving, advice, and explanations. Eliminating the gap between domain knowledge and the creation of knowledge base is still a vigorously studied issue. A clinical decision support system will improve greatly if it can be easily maintained by medical experts themselves.
Specific purposes of the system: HEALS works in a web-based environment, so it can provide services beyond the boundary of territory. In addition to the enormous general health examination applications, we also integrate healthcare domain knowledge into a highly specialized decision support system dedicated for health examinations in automatic medical data interpretation, summary, and suggestion. The simple and clear rule syntax not only reduces the learning curve to maintain the knowledge base, but also makes complex inference feasible. Additionally, the efficient reasoning algorithm gives the results in real-time. All efforts aim to improve the quality and productivity of healthcare.
Service: HEALS provides a user-friendly, intelligent, fully functional application experience to clinical workers in a distributed way. Customers can also seamlessly fetch integrated healthcare information under certain security authentication.
Evaluation results: We use health examination reports to evaluate the decision-support system of HEALS. The doctor-written final summary and suggestion of a report was compared with that of the same case produced by HEALS. The evaluation results revealed that even experienced medical experts cannot consistently perform as well as HEALS.
Conclusions: Decision-support system of HEALS is a successful evidence of interdisciplinary utilization in healthcare and information science.
Poster Proposal of 2006 American Medical Informatics Association Spring Congress