Explainable AI for PCOS Diagnosis: A YOLOv8 and Vision Transformer Approach
摘要
Polycystic Ovarian Syndrome (PCOS) is a hormonal disorder observed in individuals assigned female at birth during their reproductive age. Advancements in Deep Learning and Artificial Intelligence have led to early detection of PCOS. We have proposed a framework of YOLOv8 and Gradcam ++ integrated with the vision transformer for early detection and classification of PCOS. The most popular GradCAM ++ , an Explainable AI (XAI) is used to understand, interpret, and analyze the detection performed by YOLO V8 model. The results obtained show that the mean average precision of 0.992 and 0.995 is obtained for the PCOS and normal images and AUC of 1 with the vision transformer technique. In the future, we will focus on other popular Explainable AI for better interpretation of the models.