Risk estimation of gestational diabetes and diabetes mellitus of type -2 because of PCOD through Mathematical and Artificial Intelligence models

Abstract

Pre-existence of PCOD (polycystic ovarian disease) cause the severity of diabetes during pregnancy as gestational diabetes (GD) and post-pregnancy diabetes mellitus of type -2 (DMT-2). Early detection of PCOD may help manage the severity of diabetes mellitus in pregnancy and postnatal. This analysis conveyed to understand the pervasiveness of PCOD and its complication with diabetes mellitus and body mass index (BMI). A contextual and statistical study of the data extracted from kaggle.com in 541 patients (180 with PCOD and 361 without PCOD) of southern India has been done. The random forest (RF) technique of Artificial Intelligence (AI) model has been used to analyze the correlations among parameters. In the body mass index, 42% of 180 PCOD patients have ≥ 27 kg/m2 body mass, as waist-hip ratios are in the range of 0.80 – 1.00. With pre-existence PCOD, 35% of women are pregnant. It has observed that 84% pregnant women have the risk of developing gestational diabetes, and few women have the chance to develop diabetes mellitus of type-2. The results were analyzed by RF technique of AI through Karl Pearson’s coefficient of correlation. The patients struggling with PCOD, facing high BMI and high waist-hip ratio (0.80 – 1.00) risk of gestational diabetes, thereof early detection and diagnosis of PCOD will reduce the risk of development of GD and DMT-2.