Medical Image Segmentation Based On U-Net: A Review Of Theory And Applications

Abstract

 Medical image analysis is a process utilized by medical imaging systems to diagnose different conditions. It uses deep learning technology to analyze and improve the quality of medical images. The U-Net framework is a widely used component for semantic classification and analysis of medical images. It can efficiently segment various features of an image and provides an unbiased view of its classification. As a result, this paper offers a review of the literature on medical image segmentation based on U-Net, with an emphasis on U-effective net's segmentation experience for distinct lesion areas in six medical imaging systems. This article offers a technique for integrating the original U-Net design with deep learning, as well as a strategy for enhancing the U-Net network, in addition to the most recent developments in DL.