Impact of Channel Numbers and Training Epochs on U-Net's Polyp Segmentation Performance

Medical image segmentation plays an extremely crucial role in the modern field of medicine. It is a process that involves the precise delineation and labeling of medical images through computer technology. However, during the training of segmentation models, insufficient computational power is often a challenge. Despite having superior models, the lack of hardware resources may hinder the widespread implementation of these models. Therefore, in this paper, the U-Net model is used as a case study to investigate the impact of the number of channels and training epochs on the segmentation performance. The experimental results show that by reducing the number of channels in the U-Net model and increasing the number of training epochs, the model's parameter counts can be successfully decreased, while also achieving satisfactory segmentation results.