Association Rule Mining-based Analysis of Clinical Manifestations of SARS-Cov-2 (COVID-19) Coronavirus Infection in Senegal

The analysis of associated symptoms is a widely-studied issue in medical research, and has led to a better understanding of the clinical manifestations of several diseases such as cardiovascular disease, lung cancer, infectious diseases, type 2 diabetes, and so on. It has also been used in some studies of Covid-19 symptoms, but these have focused on patients living in Europe, America or Asia. However, as SARS-Cov-2 has claimed far fewer victims in Africa, and is unlikely to manifest in the same way on the African continent, it is important to carry out a study of confirmed cases in African countries in order to analyze the continent's specific features. In this study, we used a data analysis method known as association rule mining. This enabled us to quantify the relationships between correlated symptoms. In this paper, we used data from Senegal as the basis for experimenting our approach. The results obtained showed that the symptom sets that frequently appear in Senegal are not as broad as those noted in patients on other continents, and show specificity. This new knowledge could make it possible to define a Covid-19 clinical features specific to Senegal, and thus facilitate diagnosis and management of patients.