Preliminary Detection of Cyst Formation in Ovary Using Advanced Computer Vision Approach
Neha Ragit, Pankaj Zanke, Manasvi Jaipurkar, Rahul Agrawal, Mahima Gaurihar, Chetan Dhule · 2024
PCOS, which is a complicated endocrine disorder, severely impacts women worldwide. It is important to detect this polycystic ovary syndrome early with the objective of mitigating the risk of long-term complications. Through an emphasis on applying YOLOv8 for the analysis of ultrasound images, this study offers a novel approach for the prediction and detection of PCOS using progressive deep learning methodologies. This research showcases the development and application of a predictive model leveraging YOLOv8’s advanced computer vision capabilities to accurately detect ultrasound images indicating PCOS. Furthermore, it presents an interactive application powered by the Flask framework, enabling the real-time prediction of PCOS based on sonographic images of ovarian anatomy. Notably, the proposed YOLOv8based model has recorded an exceptional accuracy rate of $\mathbf{99.5\%}$, underscoring its effectiveness in early PCOS detection. This combination of leading-edge deep learning frameworks signifies a major achievement in PCOS diagnosis and treatment. The YOLOv8-based approach offers unparalleled accuracy and efficiency for detecting PCOS, promising transformative outcomes in healthcare for individuals affected by it.