7th ISF
COMP024 - Artificial Intelligence for Predicting Heart Disease
An electrocardiogram (ECG) is a very popular way to detect heart abnormalities and the risk of developing different types of heart disease. It is based on the principle of measuring the electrical signal generated by the heart when the heart is contracting and relaxing. The ECG can tells about the heart activities, such as heart rate or cardiac rhythm. Each segment of the ECG graph shows the relationship between the atrial and ventricular chambers. It can also help diagnose many diseases. However, detecting abnormalities with this method requires knowledge and expertise from experts in this field that sometimes may not be enough and take too much time. Also, sometimes the diagnosis can be wrong. Our project has an idea to use artificial intelligence to predict heart disease. We roughly divided the types of heart disease into ECG of normal people, ECG of supraventricular tachycardia, ECG of ventricular tachycardia, fusion graph of ECG and unknown ECG then we create our model by using neural network to learn each pattern of ECG type. Our model’s accuracy, sensitivity, specificity and f1 scores are 0.98, 0.97, 0.99, 0.97 and 0.98. Then, design a heart disease prediction program from ECG graph using this model
Show More