بررسی میزان تأثیر داروهای درمان ناباروری در بیماران نابارور با استفاده از الگوریتم خوشه بندی و تکنیک های داده کاوی

Background and purpose: The rate of infertility has increased throughout the world. Data mining is a new method for analyzing information from databases. Few studies are done regarding infertility and using data mining in describing and predicting different treatment methods and factors influencing these methods. This paper proposes a model for evaluating the efficacy of different drugs in treatment of infertility among patients treated with IUI. Material and Methods: The records of 26,035 infertile patients (from 1998 to 2009) in Sarim Hospital have been examined. Clinical data of patients were analyzed through data mining methods (Clementine V.12.0). To identify the factors influencing the efficacy of drugs classified data mining and clustering algorithms were used Results: We identified the characteristics of patients with successful treatment using K-Means clustering. CHAID decision tree helped to indicate the result of different drugs in infertility treatments. The proposed model can predict the result of used drugs with 71 percent accuracy. Conclusion: : Data mining techniques can improve the process of treatment in infertile patients by detecting the factors affecting the course of treatment.