Poster No:
2374
Submission Type:
Abstract Submission
Authors:
Sunhyun Min1,2, Younghwa Cha1,2, Marcus Pearce3, Joon-Young Moon1,2
Institutions:
1Center for Neuroscience Imaging Research, Institute for Basic Science (IBS), Suwon, Republic of Korea, 2Sungkyunkwan University (SKKU), Suwon, Republic of Korea, 3Queen Mary University of London, United Kingdom, London
First Author:
Sunhyun Min
Center for Neuroscience Imaging Research, Institute for Basic Science (IBS)|Sungkyunkwan University (SKKU)
Suwon, Republic of Korea|Suwon, Republic of Korea
Co-Author(s):
Younghwa Cha
Center for Neuroscience Imaging Research, Institute for Basic Science (IBS)|Sungkyunkwan University (SKKU)
Suwon, Republic of Korea|Suwon, Republic of Korea
Marcus Pearce
Queen Mary University of London
United Kingdom, London
Joon-Young Moon
Center for Neuroscience Imaging Research, Institute for Basic Science (IBS)|Sungkyunkwan University (SKKU)
Suwon, Republic of Korea|Suwon, Republic of Korea
Introduction:
People have diverse preferences for music, yet there are also underlying commonalities. Previous studies have suggested that complexity could be a key to musical preference. Complexity can be quantified as Information Content (IC) and entropy, as measured by the Information Dynamics of Music (IDyOM) algorithm [1]. Interestingly, an inverted U-shaped relationship has been observed between complexity and liking ratings among general listeners. This observation raises several questions: Do experts have similar preferences as compared to general listeners? And if so, do preferences correlate with distinct brain responses? To investigate, we conducted simultaneous EEG-fMRI recordings while participants listened to various music samples and collected their rating data. Specifically, we measured the relative phase of the EEG response for each piece of music and compared the results with low, mid, and high IC across the expert and non-expert groups. Through this study, we aim to deepen the understanding of how musical preferences are formed and vary among different listener groups.
Methods:
N>6 participants listened to classical music melodies inside the 3T fMRI scanner for 90 seconds each. Simultaneous EEG-fMRI measurement was carried out utilizing 64-channel EEG recording. These stimuli were generated using MIDI with the same tempo and the violin timbre of Joshua Bell (Kontakt 5). Following each piece, participants rated its complexity, likability, emotional content, familiarity, and attention using a 7-point Likert scale. Outside the scanner, participants responded to surveys including the Goldsmiths Musical Sophistication Index (Gold-MSI).
For data analysis, the expertise of music for each participant was assessed based on their scores of Gold-MSI. IC of each stimulus was calculated using IDyOM. We explored correlations between IC, participants' ratings, and neural responses. For analyzing brain responses, we utilized a method focusing on relative phase. The relative phase was defined by subtracting the global mean phase from each electrode's phase across the whole brain area. Relative phase thus demonstrates the phase-lead/lag relationships of EEG signals at each time point, revealing temporal dynamics in the brain [2][3]. We also applied relative phase analysis to the eyes closed and eyes open resting state as the references to compare. We then applied the K-means clustering method to categorize the topographic patterns of the relative phase across the whole brain.
Results:
From the ratings performed by each participant, we found a robust correlation between the perceived complexity and IC of each song regardless of their music expertise (Pearson correlation>0.7 with p-value<0.05). Also, we found an inverted U-shaped pattern when the likability and IC of each song were compared: songs with intermediate IC values were rated as most likable.
In the relative phase analysis, we observe a robust switching pattern in the brain networks between top-down mode (where the parietal regions phase-lead the frontal regions) and bottom-up mode (where the frontal regions phase-lead parietal regions). There were also transient modes where either the left or right hemisphere was phase-leading against its counter-hemisphere. The dwell time in each mode was universally highest in the eyes closed state (~230ms), followed by the eyes open state (~180ms), which suggests that external stimuli trigger mode switching. Dwell time with the music stimuli was the shortest (~160ms), supporting our hypothesis that stimuli with more content will trigger more rapid mode switching.
Conclusions:
Altogether, we found that participants, regardless of their music expertise, were able to identify the complexity of each music stimulus as identified by IDyOM. Also, there was a golden zone of IC for the music, which was most favored by the participants. Finally, our analysis showed that stimuli with more content triggered more rapid switching of the brain mode, as defined by the relative phases.
Higher Cognitive Functions:
Music 2
Novel Imaging Acquisition Methods:
EEG 1
Perception, Attention and Motor Behavior:
Perception: Auditory/ Vestibular
Keywords:
Electroencephaolography (EEG)
Perception
Other - Music
1|2Indicates the priority used for review
Provide references using author date format
1. Gold, B. P. et al. (2019), 'Predictability and Uncertainty in the Pleasure of Music: A Reward for Learning?', Journal of Neuroscience. 39 (47) 9397-9409.
2. Moon, J.-Y. et al. (2015), 'General relationship of global topology, local dynamics, and directionality in large-scale brain networks', PLoS Computational Biology, 11(4) e1004225.
3. Moon, J.-Y. et al. (2017), 'Structure shapes dynamics and directionality in diverse brain networks: mathematical principles and empirical confirmation in three species', Scientific Reports, 7(1), 46606.