Predictive Analytics And Health Monitoring System For Early Detection Of Infertility Risks Among Working Women

Abstract

Infertility on the part of working women between the ages of 22 and 30 is becoming an increasingly prevalent public health concern in urban and economically developed locations. The purpose of this study is to get an understanding of the numerous variables that contribute to infertility by analysing datasets found in the NFHS-5 and DLHS-4, as well as literature that has been examined by experts. A number of variables, including but not limited to work stress, a sedentary lifestyle, hormone imbalances, reproductive illnesses such as polycystic ovary syndrome (PCOS), and delayed marriage, are included in this category. We built a predictive analytics model employing state-of-the-art machine learning algorithms as SVM, Random Forest, Logistic Regression, Naive Bayes, and Logistic Regression to improve the early detection of infertility risk. With a forecast success rate of 93%, the Random Forest algorithm outperformed the others. Using these numbers, we can create a thorough health monitoring system that suggests all women over the age of 22 should be checked every six months. These examinations include, among other things, evaluations of the patient's mental health, testing of thyroid function, ultrasounds of the pelvic, and hormone tests (including AMH, LH, and FSH). This approach makes it easier for medical professionals to intervene at an earlier stage and improves reproductive health by providing individualised risk assessments and treatment recommendations. With the potential to improve diagnostic timings and encourage informed, timely treatment decisions, the study presents a novel approach to incorporate predictive analytics into reproductive healthcare for working women. This approach has the potential to improve diagnostic timings.