The electrocardiogram is themost accurate means of detecting arrhythmias in the heart because it recordselectrical phenomena from the heart that are not visible , touched and heard . Readingthe ECG requires careful consideration and expertise . In addition, theaccuracy of the diagnosis of arrhythmia is not high due to direct reading of thecardiac arrhythmia . In this paper, using discrete wavelet transform and theECG signal configuration window , the input of the heart is calculated afterfinding the distance between the two peaks R – R.Then all the sensitive points in the ECGthat include P – Q – R – S – T .

In this paper, the extraction of P-wavefeatures of the PR-complex of the QRS spacing of the distance between the QTspacing and the T-signal of the ECG signal using the discrete wavelet transformmethod is performed.  Afterextracting desired features, artificial neural networks have been used toclassify cardiac arrhythmias. This paper uses the PHYSIONET data base . Keywords: Electrocardiogram signal ; Cardiac arrhythmias ; PQRSTwave ; Artificial Neural Network ; ECG ; DWT ; CWT 1. IntroductionDetection of cardiacarrhythmias The ECG signal for ECG is important for diagnosis duringcardiovascular conditions. Manual analysis for the diagnosis of arrhythmias isa considerable time .2.

Methodology or SimulationMethodMethods should be described with sufficient details to allow others toreplicate and build on published results . Please note that publication of yourmanuscript implicates that you must make all materials , data , computer code ,and protocols associated with the publication available to readers. Pleasedisclose at the submission stage any restrictions on the availability ofmaterials or information . New methods and protocols should be described indetail while well-established methods can be briefly described andappropriately cited . In addition, manual analysis is always prone to error.For this reason, over the past two decades ,considerable research has been done on the automatic diagnosis of cardiacarrhythmias . The methods already presented are different from each other onhow the features are extracted, as well as the type of classification systemused.

  This paper presents a method for classifying cardiacarrhythmias using wavelet transforms and neural networks .  First , discrete wavelettransform DWT processing and extraction of ECG records the time-frequencyfeatures used. Then the feature – basedcharacteristics or time characteristics of the ECG signal. Theresulting result is used as the final attribute for training as well as the setof a multilayer particle network ( MLP ) neural network.

Although recent algorithmshave been proposed for the diagnosis of cardiac arrhythmias, most researchershave used a limited number of records in the standard MIT-BIH database.3. Natural NeuralNetwork structure  The following figure shows the structure of aneuron neuron : Each neuron consists of three main parts :* Cell body ( thetask of providing energy for the activity of the neuron )* Dendrites ( receivingelectrical signals )* Axon (transmitted electrochemical signals received fromthe cell nucleus to other neurons )