Deep Learning Based Fine-Tuned Unet for Polyp Segmentation

The identification of polyp region plays a huge role in planning and treatment of colorectal cancer. Segmentation of polyp images plays a vital role in Colonoscopy planning and implementation. The proposed framework uses traditional deep learning based fine-tuned Unet segmentation to segment and identify polyp images. An accuracy of 96.44% is obtained in segmentation of polyp images using Adam Optimizer.