Artificial Intelligence in Medical Diagnosis: Methods, Algorithms and Applications

Abstract

Artificial intelligence (AI) has evolved rapidly since the late 1980s. Increasing of healthcare datasets and its performance, the past two decades have seen an exponential progress in publications on AI. However, with the advent of increased computational power, availability of AI devices was increased. There are two main devices in AI, machine learning, where structured data (i.e. images, EP and genetic data) are analyzed and natural language processing, where unstructured data are analyzed. Both AI devices have been improved in great detail over the past two decades for its methods, algorithms, and applications. However, various attempts and new methods of AI have been used in recent years and few diseases such as cancer, nervous system disease, cardiovascular disease, liver disease, congenital cataract disease, etc. were potentially analyzed using AI. Now a day an advanced method called deep learning has initiated a boom of AI and great modifications of diagnostic medical imaging systems like endoscopic diagnosis, pathology and dermatology will be expected in the near future. Herein, the authors give a basic technical knowledge about popular methods, algorithms and applications in medical diagnosis which emerged in the past years.