Short Review on Contrastive Learning-based Segmentation Techniques for Medical Image Processing

Abstract

Due to more advancements in deep learning approaches, medical image analysis has become more popular in research. Image segmentation plays an indispensable role in image processing. Digital images are classified as segments, and segmentation approaches are used to analyze the important features and data presented in the input digital images. Segmentation is mainly performed to recover the essential features easily from the region of interest. The segmentation process generates a meaningful digital image, and they are easy to analyze. Recently, segmentation approaches have become more popular in a medical environment, and also they secured more numbers of successful applications in neutrosopy. Therefore, it is decided to make the comparative analysis of the medical image segmentation techniques based on the deep learning concept. This survey encloses various existing contrastive learning-based segmentation techniques for performing algorithmic classification in the medical domain. These surveys also compare different performance measures, datasets utilized, and tools used for the implementation. Then, upcoming research and current research gaps in medical image segmentation are analyzed. This review on state-of-the-art medical image segmentation tools has shown their potential in clinical practices for effectively diagnosing diseases with better segmentation approaches using contrastive learning.