Building the pelvic endometriosis knowledge base software

Abstract

The objective of this study is to develop a software-based medical expert system supporting the diagnosis of pelvic endometriosis. This system was developed to facilitate the creation of knowledge and inference engine. The diagnostic process used the interactive backward chaining inference algorithm. The medical knowledge data base was represented as production rules which represented in tree structures. The system was designed to interact with users in question information format. The clinical data from medical records of Gynecological out-patient clinic at HRH Maha Chakri Sirindhorn Medical Center were applied to the system by physician retrospectively. In this study, 35 medical records of women diagnosed with pelvic endometriosis were reviewed. The three most common presenting symptoms were dysmenorrhea, chronic pelvic pain and infertility, respectively. All of the patients were investigated with transvaginal sonography. Twenty-one patients had no histological studies. The clinical data of 30 patients accounted for 85.7 % were recorded successfully to the medical expert system. The diagnosis of these patients from the system corresponded with the previous data from the medical records of established pelvic endometriosis. Taken together, these data suggest that this medical expert system is a good tool to facilitate the decision making process in the diagnosis of pelvic endometriosis.