|
nif:isString
|
-
The current study was approved by the Academic Committee of the School of Psychology at South China Normal University. All participants gave written informed consent before participating in the experiments.
Eighteen volunteers (male; 18–24 years old) were recruited from local universities for the fMRI experiment. All of them were philharmonics, but not music professional, who had engaged in some musical training or activities (e.g., choir, Musical Instruments class) before the experiment. They were right handed and reported no prior history of neurological or psychiatric problems. Participants were given a small payment after the experiment.
The musical materials selected for the present research consisted of popular music excerpts, artistic music excerpts (opera), and clips of meaningless musical notes in form of playing and singing (please see supplementary material for demo, including S1 File: popular music sample, S2 File: artistic music sample and S3 File: control material sample). The former two musical materials were snipped from songs from CDs or the internet with the theme of love. These songs were all performed by female vocalists, and the lyrics were in languages other than English and Chinese. The notes clips were made in our lab and applied as control stimuli. With the GoldWave (Version 5.58, GoldWave Inc., www.goldwave.com), all stimuli were standardized to a proper length and identical volume (16dB). Each excerpt began with 500 ms of gradual fade in, and ended up with 500 ms of gradual fade out. 30 popular music, 32 artistic music and 35 notes clips were prepared for pretesting. The duration of the popular and artistic music was between 12–24 s, and that of the notes was 10–15 s. Moreover, beauty and familiarity evaluations for all these materials were collected using a 7-point Likert scale from another eighteen participants, who shared the same age and musical experience with the fMRI participants. Based on the pretesting, 40 musical excerpts were selected by matching the beauty, familiarity and length, with half for the popular music and half for the artistic music. The familarity ratings for popular and artistic music were: 3.41±0.33 vs. 3.31±0.37; the beauty ratings for popular and artistic music were: 5.17±0.35 vs. 4.98±0.37; and the lengths for popular and artistic music were: 17.80±2.65s vs. 17.85±3.88s. For these three dimensions, there was no significant difference between popular and artistic music (ps>0.05,all). 28 clips of musical notes playing and singing were chosen for the baseline condition by matching the familiarity and total duration between musical (popular and artistic music) and non-musical (note) materials. The average length of notes clips was 12.14±1.27s. The familiarity ratings for note clips (3.29±0.49) were not remarkably different from those for popular and artistic music excerpts, whereas the beauty ratings of note clips (2.81±0.33) were significantly lower than those of musical materials, both p = .000 (LSD). All the auditory stimuli were present binaurally with a high-quality MRI-compatible headphone system (SA-9800B, Shenzhen Sinorad Medical Electronics, Inc.). The volume of auditory stimuli was individually adjusted before fMRI scanning. Participants were asked to close their eyes during the experiment and open their eyes during the break. Visual instructions were presented on a screen back-projected on a head coil-mounted mirror.
The experiment program was conducted by the E-Prime 1.2 (Psychology Software Tools, Inc., Pittsburgh, PA). Prior to fMRI scanning, participants underwent a training session to become acquainted with the procedures. In the training session, the materials and the number of trials were not the same as formal experiment. The fMRI experiment consisted of 4 consecutive scanning runs. Each run contained 17 stimuli epochs, with 5 for each musical condition and 7 for the control condition. The sequences of stimuli for the four runs were presented as follows (p = popular epoch; a = artistic epoch; c = control epoch): 1) c, p, c, a, c, a, c, p, c, a, p, p, a, c, a, p, c; 2)c, p, a, c, a, p, p, a, c, p, c, a, c, a, p, a, c; 3)p, a, c, p, c, a, c, a, p, c, a, a, c, p, c, p, a, c; 4) c, a, p, c, a, c, p, c, p, a, c, a, p, c, p, a, c. Half of the participants were assigned to the main stimuli sequence (i.e., 1,2,3,4), and half were assigned to the alternative sequence (i.e., 3,4,1,2). Each run started with a blank lasting for 4 s, and then the music excerpt was presented. During inter-stimulus interval (ISI), participants had to rate the beauty of each stimulus by pressing the corresponding key. The duration of ISI was about two thirds of the presentation time of previous stimuli. The whole experimental run lasted for 420 s without any stimuli during the last 10s. All participants were asked to close their eyes during each experimental run, and open their eyes during rest. After the fMRI session, participants were required to rate the following two questions for all stimuli using a 7-point Likert scale (response format: 1 = "not at all"; 4 = "medium"; 7 = "extremely"): A. The beauty of the music; and B. The familiarity of the music.
MRI data were acquired using a 3T whole-body scanner (Siemens TIM TRIO). Functional images were obtained using a multislice echo plannar imaging (EPI) sequence (36 slices, slice thickness 3.5+0.7 mm gap, TR = 2.2 S, TE = 30ms, field of view = 220*220mm2, 64*64 matrix, flip angle: 90°). Scanning slices were aligned approximately parallel to the AP-PC plane, and interval scanning was carried out from the bottom up. For spatial normalization, a high-resolution T1-weighted anatomical image was acquired after EPI acquisition, using fast spin echo sequence(176 slices, 1×1×1mm, FOV = 256*256mm2, TE = 2.43ms, TR = 2530ms).
The obtained fMRI data were preprocessed and analyzed using the statistical parametric mapping (SPM8; Wellcome Trust Center for Imaging, London, UK; http://www.fil.ion.ucl.ac.uk/spm). For stabilization of magnetization, the first five volumes of each session were discarded. Data preprocessing was done with default setting of SPM8. EPI images were co-registered and normalized to the T1 standard template in Montreal Neurological Institute (MNI) space (resampling voxel size: 2×2×2mm), and smoothed with a Gaussian kernel with 6 mm FWHM. After preprocessing, we carried out both factorial model and parametric model analyses [13]. For the factorial model, first level analysis was performed on each subject by estimating the variance of musical epoch according to a general linear model (GLM). Three kinds of musical epochs were modeled as separate regressors convolved with the canonical hemodynamic response function. For the parametric model, trials of pop music or artistic music were respectively included into a single regressor, accompanied by a parametric regressor of according post-aesthetic rating. For both models, six motion parameters estimated during the realignment procedure were included as covariates of no interest. At the group level, all images were subjected to a voxel-wise contrast and one way ANOVA-within subject analysis to assess statistical significance for the factorial model. ROI analyses with two sample t-tests were further performed in whole brain clusters showing a significant contract between popular music and artistic music conditions. Marsbar (version 0.42) was applied to extract the beta value, with spherical ROIs of 10mm radius for putamen, 12mm radius for other areas. The central location of each ROI was determined by the results of factorial model analysis. For the parametric analysis, one-sample t-tests were used to reveal the regions in which the BOLD signal correlatively changed with the aesthetic rating scores of popular and artistic music, respectively. A tow-sample t-test was also used to find out the areas response differently to the rating of the two type of music. For the above models, global analyses were conducted at a voxel threshold of P<0.002 (uncorrected), and a cluster threshold of FDR<0.05. Small volume correction (SVC) was used, with a 10 mm radius centering a sphere on the coordinate of the ventral striatum peak voxel, and a 12 mm radius centering a sphere on the coordinate of the peak voxels of other regions. The behavioral data of scanning rating, and post-scanning ratings were analyzed by the One-way ANOVA (using LSD in post hoc analysis) via software SPSS16.
|
![[This page as RDF]](http://data.gesis.org/somesci/static/rdf-icon.gif){kind=icon}