ViT-PCOS: Vision Transformer for Automated Polycystic Ovary Syndrome Detection

This study presents Vit-PCOS, a method for detecting Polycystic Ovary Syndrome (PCOS) using Vision Transformers, aiming to act as a clinical decision support system (CDSS) for detection of PCOS from ultrasound scans, particularly in underserved areas. This study demonstrates the effectiveness of Vision Transformers (ViT) and indicates the potential to enhance diagnostic accessibility. We observe that ViT outperforms existing methods, by effectively learning discriminative features during transfer learning, and achieving a 100 percent accuracy on our test set. This illustrates the potential of utilizing Vision Transformer (ViT) in a live clinical decision support system to detect PCOS.