Artificial neural networks (ANN) are useful for function approximation and classification problems also, it has been used in many areas such as medical diagnosis, pattern recognition and so on, Electrocardiogram (ECG) is well known in medical diagnosis processes, but it used to diagnosis heart diseases.In this work, modeling is done by using ANN. The model feed forward configuration structure of two layers is used. The neural network trained using back propagation algorithm and taking 144 instances from the National Centre for Cardiology used registered by the Arduino circle. The results in the training phase were better than in test stage.The testing result show the accuracy of the system register 95% in diagnosing patients with heart diseases for National Centre for Cardiology. KEYWORDS: Neural Network, back propagation algorithm, Classification, heart diseases, Electrocardiogram, Arduino.