Advanced Neural Network Architectures for Image Processing in Medical Diagnostics

Abstract

The applications of higher-level neural network structures in medical diagnosis have significantly changed image analysis in medicamente, providing unmatched accuracy and speed of disease diagnosis and classification. This chapter provides updated knowledge of advanced neural network models, CNN, GAN, U-net models are best suited for different imaging like MRI, CT, X-ray. The findings of a literature survey are presented with focus on major contributions and open issues such as data confidentiality, model understanding, and the lack of big annotated collections. Introducing various methodological approaches for tuning neural networks for medical diagnostics, an assessment is provided of the compliance of their performance indicators with critical parameters. Knowledge and perspectives discussed in this chapter help to enhance knowledge of a reader and reveal opportunities of the artificial intelligence application for the further enhancement of the healthcare sector through enhanced image processing.