Detection of heart arrhythmia by using dual-wavelength PPG signals

Photoplethysmogram (PPG) signals are the signals obtained as a result of optical measurement of volumetric changes in the capillaries. The volumetric changes in capillaries are also dependent on the activation of the heart. As a result of recent investigations, it has been found that PPG signals contain a lot of information about the physiological and biological state of the person. Most of these studies are based on examining the waveforms of the characteristics of time and frequency spaces of PPG signals obtained by single wavelength. Electrocardiogram (ECG) signals are signals obtained as a result of electrical activity of the heart. ECG signals used in the detection of many cardiac arrhythms, including heart arrhythmias, are obtained by adhering electrodes to the body under the supervision of a health personnel in the hospital environment. Obviously, obtaining ECG signals is a laborious activity. However, the acquisition of PPG signals can be easily integrated into easy and wearable technologies. In this thesis, a new approach to the detection of cardiac arrhythmia is presented using dual-wavelength PPG signals. With the help of the hardware, attributes were extracted by using time differences between the peaks of dual-wavelength PPG signals taken from people with heart arrhythmia and healthy persons. A high degree of accuracy has been achieved in the classification made by these extracted features. This demonstrates how powerful the time difference feature is in detecting heart arrhythmia.