Eigen Decomposition of Cardiac Synchronous EEGs for Investigation of Neural Effects of Tempo and Cognition of Songs

There is growing evidence of palliative effects of listening to songs on neural and cardiovascular function. It is also known that listening to songs can entrain cardiac variability. These results suggest that the neural changes in response to listening to songs in turn affect cardiac rhythm. How these effects come about is less clearly known. Therefore, investigation of the changes in neural rhythms that are synchronous with cardiac rhythm is likely to shed further light on the mechanisms via which songs produce these effects. Towards this aim, we conducted eigen decomposition of cardiac-synchronized EEGs to investigate the effects of tempo and cognition by auditory stimuli (listening to songs). For evaluating the effects of tempo, songs of slow and fast tempo were used, and for cognition, each subjects' favorite song was used. ECG and six EEGs (F3, F4, T3, T4, P3, P4) were recorded as subjects listened to songs. For cardiac synchronization, R waves from the ECG were localized and the EEGs during 300-millisecond segments ending at each R wave were extracted. Eigen decomposition of the covariance matrix of these EEG segments was performed. Results from 14 subject showed that, compared with other locations, P3 appears to have the ability to discriminate between songs. All songs lowered the second and the third largest eigenvalues compared to control, among these, the slow tempo song induced more significant decreases in T3, T4 and P3. During the slow song, 80% of the variance in all six EEGs could be represented with less eigenvalue/vectors while during the favorite song this number was larger. These results show that eigen decomposition of cardiac synchronized EEGs has the potential to investigate effects of music on neural and cardiovascular systems.