Diagnosis of Coronary Artery Disease using Cuckoo Search and genetic algorithm in single photon emision computed tomography images

Abstract

Coronary Artery Disease (CAD) is a kind of cardiovascular disease and a heart attack is the first sign of CAD. Cardiac SPECT is one of the efficient methods to diagnose the disease. Plaque buildup in the walls of the arteries causes CAD and makes them narrow over time. Therefore, one of the most important issues is automating of CAD early detection. In the literature, various classification methods have been presented. Also, a lot of feature selection techniques have been developed to reduce the high dimension of extracted features of images in SPECT. In this paper, a method has been proposed for early diagnosis of CAD from SPECT heart images. The Cuckoo Search and Genetic algorithm are employed for selecting the optimal set of features which can lessen feature vector dimension from 44 to 5 features. Detection rate of 77.19% is obtained by using Bagging algorithm for classifying SPECT data. Results show the proposed method has high performance comparing with other recently researches.