Artificial intelligence in medical diagnostics: A review from a South African context

Abstract

Historically, South Africa, a developing country, has battled numerous chronic diseases such as Cancer, Diabetes, and Tuberculosis. Even though significant efforts are made in the medical diagnostic industry to detect and treat these chronic diseases, these efforts have fallen short due to their higher diagnostic costs, shortage in infrastructure, equipment, and highly skilled technicians at the required time, resulting in reduced access to healthcare for patients. In recent years, the field of Artificial Intelligence (AI)-based medical diagnosis has gained prominence because of its low cost, less infrastructure, equipment, and technician requirements. In addition, AI-based medical diagnosis reduces diagnostic time with a significantly high level of accuracy. Therefore, in this article, we conduct a systematic literature review of 32 collated AI articles. We, present our findings, and scope the tools, techniques, and algorithms from a South African Context. The scope of this literature review involves (1) conducting an attribute analysis of literature that includes studying disease, temporal, and spatial aspects of literature as well as stages in developing AI-based medical diagnosis tool; (2) conducting a conceptual analysis of literature that includes studying applications, algorithms, techniques, and performance measures related to different AI stages; and (3) scoping the insights from the literature from a South African context that involves proposing a framework for developing AI-based medical diagnostic tool, hardware and software requirements, and deployment strategies into underdeveloped medical diagnostic provinces of South Africa.

Keywords

• Artificial intelligence tools
• Deep learning
• Machine learning
• Medical Diagnosis