A Novel Survey on Medical Image Segmentation and Reconstruction Approaches based on Deep Learning Methodologies

Medical image segmentation or image reconstruction targets to attain the finest quality of images for clinical examination in cost-efficient and reduced patient risk. With the advent of learning approaches, deep learning and its medical imaging applications have recently attracted huge interest. This comprehensive review shows extensive analysis of various imaging techniques adopted in medical applications. This review includes an analysis of deep learning-based image construction, image reconstruction, tools used for image reconstruction and various imaging techniques. This analysis shows that deep learning- based segmentation and reconstruction approaches enhance image quality quantitatively and qualitatively. Moreover, deep learning approaches are usually cost-effectual and handle many training samples with better robustness and generalization. The basic research challenge relies on the dataset training process, which has to be addressed or resolved using modern learning techniques.