Nonlinear analysis of electrodermal activity signals for healthy subjects and patients with copd

It is known that signals recorded from physiological systems represent nonlinear features. Several recent studies report that quantitative information about signal complexity is obtained by using nonlinear analysis algorihms. Chronic obstructive pulmonary disease (COPD) is one of the causes of mortality worldwide with an increasing prevalence. This study aims to investigate different nonlinear parameters as Lyapunov exponent and correlation dimension of electrodermal activity signals recorded from healthy subjects and patients with COPD. Electrodermal activity signals recorded from 6 healthy subjects and 24 patients with COPD were analysed. Auditory, tactile and deep breathing stimuli were applied at different time intervals during the recording process. Before the analysis denoising process was applied to the signals. Signals were reconstructed in the phase space compatible with theory and largest Lyapunov exponent and correlation dimension values were calculated. Furthermore EDA and ECG signals which were recorded simultaneously from the subjects were investigated by using continuous wavelet analysis with respect to energy density and correlation between them. It has seen that for the patients with COPD the chaoticity increases with the increase in the grading of COPD. It was determined that systematic auditory stimuli increases chaoticity more than random auditory stimuli. Furthermore it was observed that participants develop habituation to the same auditory stimuli in time. Different results were found for the application of tactile stimuli to the right or left ear. The results revealed that the nonlinear analysis of physiological data can be used for the development of new strategies for the diagnosis of chronic diseases.