Author : Tabreer T. Hasan, Manal H. Jasim and Ivan A. Hashim

The area of medical information has advanced around organizing, preparing, storing, and transmit medical data for an assortment of purposes. One of these intentions is to create choice emotionally supportive networks that upgrade the human ability to analyze, treat, and evaluate forecasts of pathologic conditions. In this paper, heart disease diagnosis system has been built to classify two cases of heart conditions (Normal, Abnormal) in additional to classify five cases namely (Coronary Heart Disease, Angina Pectoris, Congestive Heart Failure, Arrhythmias, And Normal case), with high probability of classification. The proposed Heart disease diagnostic system consists of two types of database are used in the classification process; The online database which is taking from UCI learning data set repository for diagnosis heart disease and collected database from Ibn Al-Bitar Hospital Cardia Surgery and Baghdad Medical City. These databases consist of thirteen medical factors that are successful to diagnosis heart disease. Two heart diseases classifiers are proposed. They are; Multi-Layer Perceptron neural network (MLP), and Support Vector Machin (SVM). The simulation results show that, the MLP classifier has 98% accuracy of two heart diseases classification when the performance of this classifier was evaluated using collected database. While the accuracy of SVM classifier is reached 96%. Also, MLP has overcome from SVM classifier when classify four type of heart disease in additional to normal case for accuracy reached to 81%.

Keywords: Heart disease, Multi-Layer Perceptron neural network (MLP), and Support Vector Machin (SVM)

Article published in International Journal of Current Engineering and Technology, Vol.7, No.5 (Sept/Oct 2017)