Optimized Polycystic Ovarian Disease Prognosis and Classification Using AI-Based Computational Approaches on Multi-Modality Data
October 2024
in “
BMC Medical Informatics and Decision Making
”
TLDR AI can improve early diagnosis and classification of PCOS, aiding in prevention of related health issues.
The study focuses on improving the prognosis and classification of Polycystic Ovary Syndrome (PCOS) using AI-based computational approaches on multi-modality data. It utilizes a dataset of 541 instances with 45 clinical features, which are reduced to 17 through correlation-based feature extraction. Machine learning algorithms, including Logistic Regression, Naïve Bayes, and Support Vector Machine (SVM), are applied, with SVM achieving the highest accuracy of 94.44%. Additionally, 3,856 ultrasound images are analyzed using CNN and VGG16 transfer learning algorithms, with VGG16 achieving superior validation accuracy of 98.29%. The study highlights the potential of these AI approaches in early diagnosis and classification of PCOS, which is crucial for preventing infertility and other health issues associated with the condition.