Dianosing Heart Diseases Using ANN and GA
The heart is complex systems that reveals many clues about its condition in electrocardiogram (ECG), and is one of the most important organs in a human body.The walls of the heart contain myocardial tissues which contract to push the blood through the body. This contract occurs because of passing el...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English English |
| Published: |
2009
|
| Subjects: | |
| Online Access: | http://etd.uum.edu.my/1626/1/Ghassan_Nashat_Mohammed.pdf http://etd.uum.edu.my/1626/2/1.Ghassan_Nashat_Mohammed.pdf http://etd.uum.edu.my/1626/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The heart is complex systems that reveals many clues about its condition in electrocardiogram (ECG), and is one of the most important organs in a human body.The walls of the heart contain myocardial tissues which contract to push the blood through the body. This contract occurs because of passing electrical current in the heart muscle the electrical current can be captured and analyzed to diagnose the heart state. This operation is done by using electrocardiograph (ECG) device; this device captures the electrical signal, filters it from noise signals, and amplifies it. Then it displays the signal on the screen or prints it on the trace paper then the doctor interprets the ECG signal to diagnose the disease.This project discusses using artificial intelligent (AI) to process and analyze the ECG signal to diagnose the heart disease directly and display detailed report about the heart state by using the artificial neural network (ANN) after training it and finding the values of the connection weights using the genetic algorithm (GA) to choose the best values to the weights.The GA is qualified in enhancing the weights of the ANN since the ANN is trained using the classical algorithm (back-propagation), the genetic algorithm is used as a
co-training algorithm for enhancing the connection weights values and minimizing the error value. |
|---|