A Comprehensive Analysis of Deep Learning Frameworks for Gastrointestinal Tract Image Segmentation

A critical aspect of medical imaging for accurate diagnosis and treatment of gastrointestinal disorders. This study addresses the importance of this topic and aims to fill the existing research gap. By employing four distinct models, including CNN, UNet, LinkNet, and SegNet, we achieved exceptional results with $99 \%$ accuracy and an IoU score of 0.95. Our findings significantly contribute to the field of image analysis and segmentation, improving the accuracy and efficiency of gastrointestinal endoscopy image segmentation. This research has the potential to improve the uses of medical imaging, leading to more precise diagnosis and better patient outcomes. Overall, this work contributes to the progress of gastrointestinal endoscopy image segmentation, with potential applications to the medical imaging domain.