Identification of interaction between individuals by heart ratePT
Author
Lucas Marques Gasparino
Supervisor
André Fujita, Professor Ph.D
Abstract
|
Lucas Marques Gasparino. Identification of interaction between individuals by heart rate.
|
Introduction: One of the main problems in Neuroscience is the human-human interface building from brain signals analysis, which is done using expensive and unportable devices and be affected by the low signal-noise ratio.
Objective: The heart rate is easily collected by portable sensors which are more and more accurate, and it is associated with the Autonomic Nervous System, therefore we proposed to investigate its RR index as a potential indicator of interaction between subjects.
Methods: We collected heart rate data (RR signal) from three pairs of individuals (age: 23 ± 4 years, weight: 66 ± 12 kg, height:173 ± 10 cm) by sensor Polar H7 while they play Pong. In the experimental scenario, the pair of individuals played versus each other. In the control scenario, everyone played versus the computer. To identify the interaction between the players, we used correlation and Granger causality measures between the RR signal time series. To estimate the VAR model order, we used the Akaike Information Criterion.
Results: Using 5% significance level, we identified Granger causality in all pairs of individuals (100%) from the Experimental Group, while in the Control Group one out of three events (33%) had Granger causality between individuals HRV. We found bidirectional Granger Causality between the virtual players' positions in all events from the Experimental Group (100%) and two unidirectional in two events from the Control Group (67%).
Conclusion: We have evidence that the interaction between individuals could be identified by heart rate variability.
Keywords: Granger causality. Heart rate variability. Dependence between series. Human-human interface. RR index.