Diagnosis of Irritable Bowel Syndrome using data mining techniques

Chronic gastrointestinal disorders impose a heavy economic burden and stress on society and the health care system. Among these chronic disorders is irritable bowel syndrome. This disease can be diagnosed by careful evaluation of clinical signs, medical records and simple tests that only require a lot of time and high knowledge. On the other hand, data mining is one of the multidisciplinary fields derived from scientific fields such as statistics, mathematics, computers and artificial intelligence, whose applications are expanding in various fields such as health and treatment.

Applying data mining on medical data brings vital, valuable and effective achievements and can enhance the physician's knowledge to take the necessary measures and also speed up the diagnosis process. Therefore, in this study, the best data mining algorithm for the diagnosis of irritable bowel syndrome was determined using WEKA software. For this purpose, the data mining process using Decision Stump, Random Tree and j48 algorithms on 59 samples collected from the files of people referring to Erfan Hospital in Tehran (to collect information about patients) and questionnaires distributed among employees of a company. Software was used to collect information from healthy individuals. Finally, the total data included 55.93% male sample and 44.06% female sample. The results were compared in terms of speed, accuracy and computational cost. Finally, j48 algorithm with 96.6% accuracy was introduced as the optimal algorithm for IBS detection.