Coronary Artery Disease Diagnosis Using Feature Selection Based Hybrid Extreme Learning Machine

Coronary artery disease (CAD) is the most common cardiovascular disease (CVD) that cause millions of deaths worldwide due to heart failure, heart attack, and angina. The symptoms of the CAD do not appear in the early stage of the disease and it causes deadly conditions; therefore, accurate and early diagnosis of CAD is necessary to take appropriate and timely action for preventing or minimizing such conditions. Angiography, being the most accurate method for diagnosis of CAD, is often used by the clinicians to diagnose the CAD but this is an invasive procedure, costly, and may cause side effects. Therefore, researchers are trying to develop alternative diagnostic modalities for the efficient diagnosis of CAD. To that end, machine learning and data mining techniques have been widely employed. This paper proposes and develops hybrid Particle swarm optimization based Extreme learning machine (PSO-ELM) for diagnosis of CAD using the publicly available Z-Alizadeh sani dataset. To enhance the performance of the proposed model, a feature selection algorithm, namely Fisher, is used to find more discriminative feature subset. In the training period, the PSO algorithm is used to calibrate the ELM input weights and hidden biases. Further, the performance of the proposed model is compared with the basic ELM in terms of accuracy, Pearson correlation coefficient (R2) and Root mean square error (RMSE) goodness-of-fit functions. The results show that the performance of the proposed model is better than the basic ELM. The obtained CAD classification performance in terms of sensitivity, accuracy, specificity, and F1-measure is competitive to the known approaches in the literature.