Moderating effects of music on resting state networks

Abstract

Resting state networks (RSNs) are spontaneous, synchronous, low-frequency oscillations observed in the brains of subjects who are awake but at rest. A particular RSN called the default mode network (DMN) has been shown to exhibit changes associated with neurological disorders such as temporal lobe epilepsy or Alzheimer's disease. Previous studies have also found that differing experimental conditions such as eyes-open versus eyes-closed can produce measurable changes in the DMN. These condition-associated changes have the potential of confounding the measurements of changes in RSNs related to or caused by disease state(s). In this study, we use fMRI measurements of resting-state connectivity paired with EEG measurements of alpha rhythm and employ independent component analysis, undirected graphs of partial spectral coherence, and spatiotemporal regression to investigate the effect of music-listening on RSNs and the DMN in particular. We observed similar patterns of DMN connectivity in subjects who were listening to music compared with those who were not, with a trend toward a more introspective pattern of resting-state connectivity during music-listening. We conclude that music-listening is a valid condition under which the DMN can be studied.

Figure 1

Similar spatial maps for independent components (ICs) in the resting state in the music (top, n=20 subjects) and control (bottom, n=20) groups. Background is the averaged (n=40) anatomical brain. Slices are in radiological orientation (left = right). Left component (IC1): Brodmann area (BA) 9, 10, 23, and 31. Right component (IC2): BA 8, 9, 23, 30, 31, and 32 (both) + 35 and 36 (music) and 10 (control). Clusters of > 50 voxels with z-values > 2.58 are shown. Talairach Z coordinates of first and last slices of each IC are indicated. Slices are 5 mm apart.

Figure 2

Spatial maps of independent components (ICs) used as nodes and their locations or Brodmann areas (BA). Background is the averaged (n=40) anatomical brain. Slices are in radiological orientation (left = right). Left to right, top to bottom, ICs are: periaqueductal grey (PAG), thalamus (TH), basal ganglia (BG), insula, cuneus, auditory, posterior cingulate (PCC), anterior cingulate (ACC), frontopolar, and motor cortices. Clusters of > 50 voxels with z-values > 2.58 are shown. Talairach coordinates of slices are indicated.

Figure 3

Undirected graph of partial spectral coherences (PSC) between nodes from Figure 2. Midline and bilateral nodes are shown over the left hemispheric surface. Edges are drawn between nodes with PSC > 0.3261 inflection point. Locations or Brodmann areas (BA) of nodes are indicated. PAG = periaqueductal grey, TH = thalamus, BG = basal ganglia. Nodes identified as part of the canonical default mode network (DMN) are colored gold. All other nodes are colored green. Edges colored red are present only in the music group, edges colored blue are present only in the control group, and edges colored purple are present in both groups. [Background image from Wikimedia Commons: Figure 728 from 1918 Gray’s Anatomy vectorized by user Mysid.]

Figure 4

Figure 4A (top two rows): Spatial map of independent component (IC) corresponding to the default mode network (DMN) colored green. Voxels with intensity greater than 20% of the maximum are shown. Background is MNI152 T1 1mm brain. Slices are in radiological orientation (left=right). MNI152 Z coordinates of first and last slices in the IC are indicated. Slices are 4 mm apart. Figure 4B (bottom one row): Group-level differences for the DMN IC in the corresponding slices of Figure 4A (top) obtained via spatiotemporal regression (STR). Voxels with p-values less than 0.05 (1 - p-value > 0.95) after correction for multiple comparisons are shown. Voxels with greater coactivation in the music group are colored red. Voxels with greater coactivation in the control group are colored blue. There is greater activation in right-sided BA 19 in the music group and in left-sided BA 18 and 19 in the control group (see text for more detail).