DEVELOPMENT OF DEEP LEARNING APPLICATIONS IN MEDICAL IMAGE ANALYSIS

Abstract

The big performance of machine learning algorithms at image recognition tasks in recent years intersects with the era in which electronic medical reports and diagnostic imaging are significantly increased. This study presents the algorithms for machine learning applied to medical image processing, which rely on neural networks in convolution and highlight clinical aspects. The benefit of machine learning in a medical broad data period is that important hierarchical connections inside the data may be identified algorithmically without laborious hand-made tools. We cover key research areas and applications for the classification, location, detection , segmentation and registration of medical images. Finally, we address study challenges , new developments and possible future paths.