Publication

Abstract : Polycystic Ovary Syndrome (PCOD) has emerged as a significant health concern in India, with prevalence rates reported between 9% to 36%. However, societal stigmas and misconceptions hinder the timely recognition and diagnosis of PCOD. A multi-modal AI-based framework for PCOD Detection and Risk Assessment is presented considering algorithms that include SVM, CNN, KNN, Otsu, and Näıve Bayes. This work integrates both image and numerical data utilizing ultrasound images, blood test results, and various patient parameters. 12 key parameters: height, weight, BMI, FSH, TSH, LH, AMH, PRL, Vitamin D3, progesterone, and RBS are considered. We conducted a survey involving 81 participants. Our survey revealed that 45% of participants were unaware of their PCOD status despite exhibiting symptoms, and 73% of participants exhibited symptoms of PCOD. 19.8% of respondents confirmed PCOD diagnoses. Our framework enhances diagnostic precision but also mitigates the emotional toll wrought by PCOD through early intervention and support.