Advancements In Medical Imaging Based on Image Denoising, Data Augmentation, and Classification With Computational Techniques

Abstract

Nowadays, medical imaging techniques are often not directly suitable for disease classification due to various noise factors and limited image availability. Innovative approaches to image denoising, data augmentation, and classification have been made possible by the fast development of computer technologies, which have had a significant influence on medical imaging. The use of Generative Adversarial Networks (GANs) for image denoising and the creation of synthetic data to supplement training datasets are the primary focus of this review, which analyzes the development and incorporation of state-of-the-art computational methods. This survey also highlights the critical function of GANs in separating important anatomical features from background noise, greatly improving the accuracy of medical picture analysis. Exploring data augmentation by synthetic picture synthesis also reveals how it helps overcome dataset restrictions to train deep learning models. In addition, the developments in classification approaches, this survey highlights how Convolutional Neural Networks (CNNs) and ensemble learning have revolutionized diagnostic capacities and detection accuracies in security applications. This survey lays out the complex effects of these technology developments on the accuracy and trustworthiness of medical imaging procedures by synthesizing all the latest research findings. © 2024 IEEE.