Pcos Prediction: Advancements in Medical Informatics Using Artificial Intelligence

A hormonal condition known as a polycystic ovarian syndrome (PCOS) affects around 1 in 10 persons throughout their reproductive years. Weight gain, ovarian cysts, and irregular menstruation periods are all signs of PCOS. A hormone called androgens, which may cause excessive facial hair, acne, and other cosmetic changes, is linked to the disorder via elevated amounts. In addition to higher risks for miscarriage. According to a research paper from the overall population of India, 11.8% of women are not diagnosed with PCOS and 22.7% of women are suffering from PCOS.PCOS may be accurately predicted and early identified, which can considerably improve patient outcomes and enable appropriate therapies. This study examines the use of multinomial logistic regression analysis to forecast the development of PCOS based on pertinent clinical and demographic data. An online survey was conducted to collect the data for the research. About 73.2% of women who participated in the survey said they had never tested or PCOS. The trained model had an accuracy of 82%. The mean crossvalidation for the model is 0.75%. A confusion matrix was used to assess the proposed model's classification performance. The PCOD/PCOS categorization problem's anticipated and actual classes were shown in a matrix. All 22 samples were accurately categorized into class 1 (PCOD/PCOS present), demonstrating the model's strong predictive accuracy. Twenty of the twenty samples for class 3 (missing PCOD/PCOS) were correctly categorized. Class 0 (misclassified as PCOD/PCOS missing) had two misclassifications from class 1 and one correct prediction. Finally, the classification of all four samples from class 4 (PCOD/PCOS type 5) was correct. These findings emphasize the model's potential l for clinical use by showing how well it can predict the likelihood of PCOD/PCOS.