Intelligent Detection of Heart Disorders

In this chapter, the first stage for detecting heart disorders (that is, noise removal) is explained. Two intelligent approaches based on Self Organizing Map (SOM) and Particle swarm Optimization (PSO) are used to train the feedforward neural network for noise removal. The trained ANNs are used to find the cutoff frequency. Then the found cutoff frequency is applied by a bandpass FIR filter for ECG noise removal.

Feature Extraction and Classification

In the previous chapter, the first stage for detecting the ECG noise removal was investigated. In this chapter, the second and the third stages are explained. The Second stage is to extract the effective features of the ECG signals. The final stage is to use MLP and PSO algorithms for classification of ECG signals to detect the 4 common heart disorders including the normal signals. Common disorders are Normal, Supraventricular, Brunch bundle block, Anterior myocardial infarction (Anterior MI), and Interior myocardial infarction (Interior MI).

ECG Signal Denoising Algorithm Based on Optimization Wavelet Transform

Wavelet transform algorithm with its unique multi-resolution analysis and it is in the time - frequency domain has the advantage of the ability to characterize the local signal characteristics, let it has been widely used in signal detection, noise removal, feature extraction, image compression and so on. In this paper, on the basis of already wavelet transform ECG noise removal, proposed a median filter optimization algorithm, enables ECG noise removal effect is more obvious, also for the Eigen values detection of ECG lay a better foundation.