Use of Artificial Intelligence in Cardiology: Where Are We in Africa?

Abstract

Cardiovascular diseases are the leading cause of death. Their inherent silent nature makes them often challenging to detect very early. The management of these diseases also requires many resources. Meanwhile, Artificial Intelligence (AI) in cardiology has recently showed its ability to fill the gap. Indeed, several scoring methods and prediction models have been developed to understand the different aspects of these pathologies. The purpose of this paper is to review the state-of-the-art of the use of AI and digital technologies in cardiology in developing countries and to see the place of Africa. We have conducted a bibliometric analysis with 222 papers and an in-depth study on 26 papers using real and local databases. The words arrhythmia, cardiovascular disease, deep learning, and machine learning come up most often. Support vector machine algorithms, decision tree-based assemblers, and convolutional neural networks are more used. Among the 26 papers studied, only one comes from Africa, 24 from Asia, and one is a joint work between researches from Uganda and Brazil. The results show that countries using these AI-based methods often have accessible health databases, and collaborations between health specialists and universities are frequent. The finding of the African studies is that they focused, in most instances, on medical research to find risk factors or statistics on the epidemiology of heart disease.