PET images, obtained by the^15 0-bolus method, were coregistered on each participant’s
MRI image and spatially normalized.22–27The grand-mean image for each task was com-
pared to control, forming grand-mean images of the task-induced changes in brain activ-
ity (p0.001). Grand-mean images were converted to z-score images using the population
variance.^28 Activation loci were identified by 3-D center-of-mass address, peak-voxel inten-
sity, and z-value, cluster size, and statistical significance.^29 Response coordinates were
interpreted^30 in accordance with the Talairach and Tournoux stereotaxic atlas^23 and con-
firmed with coregistered group-mean MRI.
Only a brief summary of some of the principal findings can be presented here. I will
focus on the auditory cortex (the temporal lobe). This is where we detected the primary
difference in neural activation for Bach and for scales, as indicated in the analysis of the
direct contrast between activity in the two conditions. As illustrated in Figure 17.2, many
of the activations were in the anterior secondary auditory association areas, in Brodmann’s
areas (BA) 22, 21, and 20. Overall, the performance of the Bach activated the temporal lobe
more strongly than did the performance of scales. Scales performance activated middle
temporal areas bilaterally, with more on the left than on the right (Figure 17.2). By con-
trast, the performance of the Bach activated superior, middle, and inferior temporal areas
bilaterally, with more on the right than on the left.
249Figure 17.1A pianist in a positron emission tomography (PET) scanning session^21 performing the third move-
ment of the Italian Concerto by J. S. Bach. Photograph by Dr Stephan Elleringmann (Aurora/Bilderberg).