CARDIO-RESPIRATORY SIGNAL ANALYSIS DURING HUMAN RHYTHMIC MOVEMENT

This thesis focuses on analyzing physiological signals from the cardio-respiratory system of human beings during movements. The cardio-respiratory system plays a crucial role in maintaining overall health and well-being, and analyzing its signals provides valuable insights into its functioning. However, in real-life scenarios, such as daily living and physical activity, body movement introduces distortions that affect the accuracy and interpretation of these signals.

The objective of this research is to identify and attenuate the impact of motion artifacts on the analysis and interpretation of cardio-respiratory signals, specifically heart rate variability (HRV), during human movements. The study also aims to estimate important ventilatory parameters from these signals in real-life scenarios.

The study utilizes a methodology that integrates the processing of physiological signals with practical, real-world situations. Healthy adults are monitored using devices that collect acceleration, electrocardiogram (ECG), and respiratory signals while they engage in different activities. The collected data is processed and analyzed using analytical methods (such as ECG signal processing, HRV analysis, and machine learning algorithms) to interpret the physiological information.

The findings reveal that motion artifacts significantly affect HRV during pseudo-rhythmic exercises, such as during running and cycling. Time-frequency analysis of HRV, combined with acceleration signal analysis enables the identification and quantification of body movement impact on HRV. We propose a solution based on error correction techniques that reduce the disruptive influence of motion distortions on HRV time-frequency analysis, resulting in improved accuracy and reliability.

Moreover, this study investigates the estimation of ventilatory markers in uncontrolled environments. Spectral fusion techniques are used to estimate breathing rate (BR) during everyday activities, and HRV indices are used to estimate the second ventilatory threshold (VT2) during maximal exercise tests. These estimations offer insights into the functioning of the cardio-respiratory system within real-life situations.

This research improves the interpretation of cardio-respiratory signals in real-world scenarios, addressing motion artifact distortions and providing practical mitigation strategies. These insights benefit sports physiologists and researchers, enhancing their assessments of cardiovascular health and physical fitness in individuals.