QUANTIFICATION OF THE RESPIRATORY TIME-SERIES REGULARITY AND COMPLEXITY USING APPROXIMATE ENTROPY AND SAMPLE ENTROPY

摘要

The objective of this article is an attempt to measure complexity and regularity of the physiological signals during sleep in direction of detection and classification of the sleep apnoea syndrome. We use approximate entropy (ApEn) and sample entropy (SampEn) to assess diversity of consecutive breathing patterns. Experimental investigations are preceded by the theoretical and computer analysis for the example of stochastic process MIX(P) and the cardiorespiratory PNEUMA model. Realisation of the research procedure resulted in the statement of the potential of the algorithms to distinction between normal and pathological respiratory conditions during sleep and quantitative evaluation of the observed changes. In conclusion, some important directions were outlined for the future studies.