Comparative Study of Three Artificial Intelligence Algorithms for the Prediction and Classification of Cardiovascular Diseases from ECG Signals: Application to Chadian Patients

In this paper, a comparative study of three artificial intelligence algorithms: Artificial Neural Networks (ANN), k-Nearest Neighbors (KNN), and Logistic Regression (LR) is conducted for the prediction and classification of cardiovascular diseases (CVD) based on electrocardiogram (ECG) data from Chadian patients.The experimental analysis relies on a dataset of 1,025 patients, comprising 14 clinical and electrocardiographic attributes, collected from local hospital databases. Model performance was evaluated using standard metrics such as accuracy, precision, recall (sensitivity), and F1-score. The results demonstrate that the ANN model outperforms the other approaches, achieving an accuracy of 99.6%, a precision of 100%, a recall of 99.4%, and an F1-score of 99.6%, compared to slightly lower performance obtained with the KNN and LR models. These findings confirm the strong potential of artificial neural networks for ECG signal analysis and decision support in the diagnosis of cardiovascular diseases. In the Chadian context, characterized by limited medical and technological resources, this approach represents a promising solution for early detection of CVD and improved patient management. However, further studies are required to validate the robustness of the proposed model on larger datasets and to integrate additional clinical risk factors.