DxGenerator: An Improved Differential Diagnosis Generator for Primary Care based on MetaMap and Semantic Reasoning

Background: In recent years, researchers have used many computerized interventions to reducemedical errors, the third cause of death in developed countries. One of such interventions is theuse of differential diagnosis generators in primary care, where physicians may encounter initialsymptoms without any diagnostic presuppositions. These systems generate multiple diagnoses,ranked by their likelihood; as such, the accuracy of these reports can be determined by the locationof the correct diagnosis in the list.Objective: This study aimed to design and evaluate a novel practical web-based differential diagnosisgenerator solution in primary care.Methods: In this research, a new online clinical decision support system, called DxGenerator,was designed to improve diagnostic accuracy; to this end, an attempt was made to converge a semanticdatabase with the unified medical language system (UMLS) knowledge base, using MetaMaptool and natural language processing (NLP). In this regard, ۱۲۰ diseases of gastrointestinalorgans, causing abdominal pain, were modeled into the database. After designing an inferenceengine and a pseudo-free-text interactive interface, ۱۷۲ patient vignettes were inputted into Dx-Generator and ISABEL, the most accurate similar system. The Wilcoxon signed ranked test wasused to compare the position of correct diagnoses in DxGenerator and ISABEL. The alpha levelwas defined as ۰.۰۵.Results: On a total of ۱۷۲ vignettes, the mean and standard deviation of correct diagnosis positionsimproved from ۴.۲±۵.۳ in ISABEL to ۳.۲±۳.۹ in DxGenerator. This improvement wassignificant in the subgroup of uncommon diseases (P-value < ۰.۰۵).Conclusion: Using UMLS knowledge base and MetaMap Tools can improve the accuracy ofdiagnostic systems in which terms are entered in a free text manner. Applying these new methodswill help better accept medical diagnostic systems by the medical community.