Development of a computer program for the diagnosis of polycystic ovary syndrome in women of early reproductive age

Abstract

Aim. To develop the software for the diagnosis of polycystic ovary syndrome in women of early reproductive age. Material and Methods. We consecutively recruited 200 women of early reproductive age (18-35 years) with or without polycystic ovary syndrome (100 women per group) and performed standard clinical, laboratory, and ultrasound examination. In particular, we evaluated the duration of the menstrual cycle (oligo/ amenorrhea), ovarian reserve, and ultrasound parameters of the uterus and appendages (antiMullerian hormone level, ovarian volume, and antral follicle count). Anti-Mullerian hormone level was measured by means of enzyme-linked immunosorbent assay while ovarian volume and antral follicle count were examined on the 3rd day of the menstrual cycle using a transvaginal ultrasound examination. Results . We found statistically significant differences in the indicators of ovarian reserve in women of early reproductive age between the patients with and without polycystic ovary syndrome. Employing a logistic regression analysis, we developed a software for the automated diagnosis of polycystic ovary syndrome. Sensitivity and specificity of this software were 70.9% and 75.7%, respectively. The «Clinical, Laboratory, and Ultrasound Diagnostics of Polycystic Ovary Syndrome (CLOUDPOS)» software was developed and registered by Federal Intellectual Property Service (certificate No. 2019662249, date of registration 09/19/2019). Conclusion. Our software can be recommended as an accessory tool for the diagnosis of polycystic ovary syndrome.