Review of Deep Learning in Medical Imaging

Abstract

In the past few years, computer-aided analysis has fallen quickly behind. Organ categorization and segmentation is a key stage in computer-aided diagnosis. In recent years, there has been a lot of interest in the division of abdominal organs such as the liver, stomach, kidney, pancreas, and bladder from various picture modalities. Radiologists or medical specialists analyze abdominal pictures the majority of the time. Due to subjectivity, intricacy of the picture, substantial variations between interpreters, and fatigue, human specialists' ability to understand images is quite restricted. Following the success of deep learning in real-world applications, it is now offering innovative solutions for medical imaging with high precision and is regarded as a crucial technique for upcoming medical applications. When compared to conventional models, the development of deep convolutional neural networks (CNN) tends to provide superior categorization in medical imaging analysis. Although CNN is the most frequently used architecture in medical picture analysis, this form of analysis also uses supervised machine learning methods, which need a lot of training data. A data set of images related to the abdomen can be used to test new unsupervised learning algorithms and architectures like VAE (Variational auto encoders) and GAN (Generative adversarial networks) that are emerging in the field of deep learning to get a sense of how performance measures related to diagnostic results might improve. This can be done to gain a greater understanding of how to enhance the diagnosis performance.