Your Brain on Art: Emergent Cortical Dynamics During Aesthetic Experiences

Abstract

The brain response to conceptual art was studied with mobile electroencephalography (EEG) to examine the neural basis of aesthetic experiences. In contrast to most studies of perceptual phenomena, participants were moving and thinking freely as they viewed the exhibit The Boundary of Life is Quietly Crossed by Dario Robleto at the Menil Collection-Houston. The brain activity of over 400 subjects was recorded using dry-electrode and one reference gel-based EEG systems over a period of 3 months. Here, we report initial findings based on the reference system. EEG segments corresponding to each art piece were grouped into one of three classes (complex, moderate, and baseline) based on analysis of a digital image of each piece. Time, frequency, and wavelet features extracted from EEG were used to classify patterns associated with viewing art, and ranked based on their relevance for classification. The maximum classification accuracy was 55% (chance = 33%) with delta and gamma features the most relevant for classification. Functional analysis revealed a significant increase in connection strength in localized brain networks while subjects viewed the most aesthetically pleasing art compared to viewing a blank wall. The direction of signal flow showed early recruitment of broad posterior areas followed by focal anterior activation. Significant differences in the strength of connections were also observed across age and gender. This work provides evidence that EEG, deployed on freely behaving subjects, can detect selective signal flow in neural networks, identify significant differences between subject groups, and report with greater-than-chance accuracy the complexity of a subject's visual percept of aesthetically pleasing art. Our approach, which allows acquisition of neural activity "in action and context," could lead to understanding of how the brain integrates sensory input and its ongoing internal state to produce the phenomenon which we term aesthetic experience.

Keywords: EEG; aesthetics; freely moving; functional connectivity (FC); machine learning.

Figure 1

Data processing flow chart.

Figure 2

Flow chart of clustering method for visually pleasing pieces in the exhibit.

Figure 3

Patterns of interest in functional connectivity analysis. (A) All possible connections from electrodes in the anterior region to electrodes in the posterior region were grouped together for each subject during Piece 1 viewing and baseline. (B) Posterior to anterior connections. (C) Left to right connections. (D) Right to left connections. (E) Left to left connections. (F) Right to right connections.

Figure 4

Results of the questionnaire data (n = 207). Participants were asked to select the most emotionally stimulating and aesthetically pleasing pieces. The age and gender distributions of all participants, volunteering questionnaire data for this study, are also shown.

Figure 5

Clustergrams (left) of standardized EEG signals and representative EEG records (right) from subjects viewing Piece 6 (n = 16). EEG traces at right are from channel FC6 for the indicated subjects. Average ± SD ages for two clusters (depicted as red and blue) of participants were 24.3 ± 2.4 and 31.1 ± 12.6 years, respectively. Open circles at the end of the dendrogram tips indicate the rows matching EEG traces at the right. Vertical ordering of EEG traces matches the vertical order of their corresponding clustergram rows.

Figure 6

Image analysis and clustering accuracy results. (A) Hierarchal clustering of image-based features to determine those pieces in the exhibit that shared certain visual aspects. Piece 3 was not included in the analysis since a complete image of the piece was not available. (B) Clustering accuracy results using the groupings identified in (A) for three different feature domains and two clustering methods.

Figure 7

Important channels for distinguishing between different aesthetically evoking visual art pieces according to the feature selection algorithm, mRMR. Each column represents the different feature domains used to cluster the data. The different colors represent the percentage of use of that particular channel in the selected most important 50 features for clustering. The rightmost column shows the labels of the channels.

Figure 8

Percentage of top 50 features in the time, wavelet, and frequency feature domains from each frequency band for the all studied subjects (n = 20).

Figure 9

Distribution of connectivity coefficients for predefined patterns (cf. Figure 3) using EEG data from viewing Piece 1 (blue line) and the baseline data (red line). Distributions are shown for the delta band (A–F) and gamma band (G–L). Distributions are normalized to the number of counts in each connectivity strength bin. All comparisons between piece viewing and baseline data for each pattern are statistically significant (Wilcoxon test, p < 0.05).

Figure 10

Scalp maps showing the number of connections between functionally-related pairs of electrodes. The channel that yielded the most connections (averaged over subjects) within each defined pattern was plotted on the scalp map. Results are shown for the delta and gamma bands.

Figure 11

Average connectivity coefficients for the defined patterns over 1-s time intervals in the delta band for two participants. The top row of scalp maps represents the baseline data. The bottom row of scalp maps represents the Piece 1 viewing data. The patterns that each line arrow represents are shown in the top left scalp maps. (A) The subject is a 49 y/o female. (B) The subject is 27 y/o/ male.

Figure 12

Distribution of connectivity coefficients for predefined patterns using data from all female (red line) and male (blue line) subjects viewing Piece 1. Distributions are shown for the delta band. Distributions are normalized to the number of counts in each connectivity strength bin. All comparisons between male and female for each pattern are statistically significant (Wilcoxon test, p < 0.05). In plots (A, C–F), males have significantly higher connectivity coefficients. In plot (B), females have significantly higher connectivity coefficients.

Figure 13

Comparison of the average connectivity coefficients for the two youngest subjects and the two oldest subjects viewing Piece 1. The average coefficients were calculated for each frequency band.