The aim of this project is to analyse and observe the pattern of variation of electrocardiogram components and segments with varied respiration and heart rate.
Multi-channel ECG and respiration signal are recorded simultaneously from volunteers and patients. Afterwards, Signal-to-Noise Ratio (SNR) is improved and baseline wander artifact is greatly eliminated without any morphology distortion. Applying an accurate delineation process, ECG significant points are located. Finally, a powerful multivariate statistics technique (Principal Component Analysis) is used to identify linear correlation between detected and extracted ECG components under consideration.