Article Preview
Top1. Introduction
Since 1903 the Electrocardiogram (ECG) was introduced to clinical medicine, the techniques have been developed rapidly in the record, processing and diagnosis of the ECG whether it is in the biomedical area or in engineering and we accumulate considerable experience. ECG plays an important role in the clinical diagnosis of the heart disease. It provides an objective indicator for correct analysis, diagnosis, treatment and care of the heart disease. Because of its important social value and economic value, it has a very wide range of applications in the modern medicine. ECG is still a major research subject in the biomedical engineering.
ECG is the recording of the electrical activity of the heart, and has become one of the most important tools in the diagnosis of heart diseases. ECG signal is shaped by P wave, QRS complex, and T wave. In the normal ECG beat, the main parameters including shape, duration, R-R interval and relationship between P wave, QRS complex, and T wave components are inspected. Any change in these parameters indicates an illness of the heart.
The study of ECG recognition has an important significance in understanding human heart in the role of human intelligence. Although many efforts have been taken recently to recognize ECG using different methods, current recognition systems are not yet advanced enough to be used in realistic applications.
1.1. Neurons
As we know that the human brain is a collection of Neuron’s, on an average 1011 Neuron’s are in human mind. These neurons are called Processors, on an average connections in human brain are 1000 – 10,000 connections. As we know that Mind works when Activation value is greater than threshold value. Initially Axon: - is used to transmit data to one Neuron to another.
And Synapse is the point of contact between neurons. Generally, mind is a collection of electro chemical processes, electrical waves, and the pulse generated by the neuron and that travel along the Axon as an electrical wave and a chemical wave, once these pulses reach the synapses at the end of the axon they open up a chemical vessel.
1.2. Learning
1. Supervised Learning: - Training set contains both input and required output 2. Unsupervised Learning: - It is based on clustering of input data. In this no target value is Known 3. Reinforcement Learning: - In this correct output is not known. Means we know only partially knowledge about output.
1.3. Back Propagation Algorithm
Most common methods of obtaining weights in the networks is a method of supervised learning means we know the output to corresponding input. This method is used for minimizing the error in the network.
In this paper we find out the error in the form of high frequency wavelength (in the form of ECG) of a person who is suffering from heart disease. In this article we take the ECG of a normal person who is not suffering from any disease. And we also collect the data (frequency wavelength) of those people which are suffering from heart disease, and we convert this frequency in wave form.
If a person is suffering from heart disease then his frequency must be higher from a normal person, then we compare that frequency to the original one (normal person).
If we compare and find the error by the use of back propagation algorithm. We use log sigmoid function for finding the error. And we increase or decrease the weight by using medicine which is given to patient, it depends on the mismatch of the wave length.