An Electrocardiogram Classification Based on The Walrus Optimization Algorithm Improved The BP Neural Network Intelligence Model

Heart diseases are one of the most common life-threatening factors, especially due to abnormal heart rhythms known as arrhythmias.Early detection of these arrhythmias can have a great impact on the prevention and successful treatment of heart diseases.An electrocardiogram (ECG) is a key tool for recording the electrical activity of the heart, and the automatic classification of ECG signals plays an important role in identifying arrhythmias and heart diseases.However, one of the challenges in this field is the low accuracy of heart disease prediction models and the inefficiency of feature selection methods.In this paper, a new technique named Walrus Optimization Algorithm (WaOA) in combination with Backpropagation Neural Network (BPNN) is presented to classify ECG signals.To improve the performance, wavelet transform with four different functions and adaptive thresholding techniques have been used to filter ECG signals and extract features.Then, with the help of Principal Component Analysis (PCA), the dimensions of the extracted features were reduced to minimize the computational complexity of the model.The Walrus optimization algorithm helps to improve the performance of the backpropagation neural network and allows the model to be optimized automatically.To evaluate this technique, the public MIT-BIH Arrhythmia Database has been used.Experiments have shown that the WaOA-BPNN technique achieves a high accuracy of 98.80% in the classification of ECG signals.