Dataset for the Development of AI-Powered Diagnostic Models for Vector-Borne Diseases: a perspective from Burkina Faso

Abstract

Vector-borne diseases, such as malaria, dengue, and yellow fever, pose a major public health challenge, particularly in the Hauts-Bassins region of Burkina Faso, where access to healthcare services is often limited. This dataset provides a structured collection of clinical and demographic data from 300 patients in the Dafra and Do health districts, collected over a five-week period. It is carefully annotated with patient symptoms, medical history, and laboratory-confirmed diagnoses, offering a valuable source for symptom analysis and identification of epidemiological trends. The analysis of patient records revealed that malaria was the most prevalent disease, accounting for 73.57% of cases. The majority of patients were female, representing $\mathbf{5 1. 0 2 \%}$ of the sample. Additionally, an imbalance ratio of 22.5 highlights a significant disparity in class distribution, which may lead to a biased model favoring the majority class. Furthermore, 20 symptoms were observed in at least 20% of the patients. By making this dataset available, we aim to facilitate the development of artificial intelligence models for the automatic diagnosis of vector-borne diseases. In addition, this data set serves as a valuable resource for local public health initiatives, allowing scalable disease surveillance and improving diagnostic accuracy in underserved areas.