This chapter uses intelligent methods based on swarm intelligence and artificial neural network to detect heart disorders based on electrocardiogram signals. This chapter has introduced the methodology undertaken in the denoising, feature extraction, and classification of ECG signals to four heart disorders including the normal heartbeat. It also presents denoising using intelligent methods.
Chapter Preview

7.3 Methodology Stages

A database of 100 lead II ECG signals were obtained from the Physiobank database. Figure 1 shows the simplified overall algorithm for the classification.

Figure 1.

Flowchart of overall methodology for ECG classification

Preprocessing: The noise from powerline interferences, as well as motion artefacts from the electrode and skin interface, affect the QRS complex, P and T waves of the ECG signals. In the preprocessing, two intelligent approaches based on self-organising map (SOM) and PSO neural network (PSONN) for finding the cutoff frequency, are proposed and applied with one of the traditional models commonly used by many researchers (Losada, 2004; Orfanidis, 1996; Lian and Hoo, 2006; Engin, 2004; Minami et al., 1999; Lin et al., 2006; Naghsh-Nilchi and Kadkhodamohammadi, 2008). In this project, Finite Impulse Response (FIR) filter FIR is preferred to the infinite impulse response (IIR) filter because IIR filters are more susceptible to problems of finite-length arithmetic. The baseline wander is removed using a median filter as recommended in (Poungponsri and Yu, 2009).

Complete Chapter List

Search this Book: