Book

Description

To investigate the association between heart activity and sound stimuli, a methodology based on digital signal processing (DSP) and artificial intelligence (AI) was applied in order to examine the effect of musical structures and noise on ECG and HRV. In experiments, it was possible to recognize the valence judgment of participants using binary evaluation and ECG signals. The outcomes revealed a high performance of 0.91 AUC and a similarly high accuracy of 0.86 in recognizing emotional valence, as elicited by instrumental sections of music. Using machine learning, noises and harmonic intervals were found to alter heart activity distinctively: the ratio between the axis of the ellipse fitted in the Poincaré plot was found to change on exposure to harmonic intervals and noise. Moreover, the employed techniques, DSP and AI, were able to demonstrate that the heart reacted differently to different stimuli. For example, the frequency content of the noise and harmonic intervals produced different heart responses and in the case of harmonic intervals, the effect of consonance quality could be detected in the heart response, supporting the theory concerning the biological influence of consonance-dissonance perception. It was observed that different types of noise, harmonic music intervals, and instrumental sections of music generated different patterns in the captured heart signals recognized by the techniques used in the analysis. The results of this research thus represent a valuable contribution to the fields of perception and health.