Disease Similarity Calculation on Simplified Disease Knowledge Base for Clinical Decision Support Systems

For clinical decision support systems designed to help physicians make diagnostic decisions, “disease similarity” data is highly valuable in that they allow continuous recommendation of diagnostic candidates. To build such a recommendation algorithm, this paper explores a method to measure disease similarity between diseases on a simplified disease knowledge base. Our disease knowledge base comprises disease master data, symptom master data, and disease–symptom relations that include clinical information of 1550 disorders. The calculation of the disease similarity is performed on this knowledge base, with i) standardized disease classification, ii) probabilistic calculation, and iii) machine learning, and the results are evaluated with a gold standard list audited by a physician. We also propose a novel metric for evaluation of the algorithms to calculate the disease similarity. A comparative study between the algorithms revealed that the machine learning approach outperforms the others. The results suggest that even a superficial calculation on a simplified knowledge base may satisfy the clinical needs in this problem domain.