TABLE 5.

Results of analyses between musical activity and GMV at the voxel level.

No. cluster	Label (Hammersmith atlas)	Brodmann area (BA)	Hemisphere	Cluster		Peak of cluster
				P	size	Z value	MNI coordinates (x y z)
Main effect: musical activitya
1	Postcentral gyrus	BA 6	Left	0.195	220	3.60	–57	–6	27
Interaction effectb
1	Precentral gyrus†	BA 6	Left	0.025	718	4.90	–28	–10	48
2	Precentral gyrus	BA 4	Left	0.114	322	4.81	–39	–14	33
3	Inferior middle temporal gyrus	BA 20	Right	0.054	502	4.45	51	–12	–42
4	Fusiform gyrus	BA 20	Left	0.279	145	3.99	–36	–8	–38
5	Inferior frontal gyrus	BA 46	Lateral/right	0.140	277	3.90	40	33	12
6	Anterior medial temporal lobe	BA 38	Left	0.245	168	3.78	–38	16	–36
7	Orbito-frontal lobe	BA 11	Medial/right	0.166	242	3.59	6	34	–14
8	Postcentral gyrus	BA 4	Right	0.293	137	3.52	50	–8	26

Models adjusted for scanner site and TIV.

Musical activity was included as binary predictor, dummy coded with musical activity = 1, no musical activity = 0.

^aResults from the main effect model with musical activity and GMV (p < 0.001 uncorrected, expected voxels per cluster k = 140).

^bResults from the interaction effect model with musical activity, working memory, and GMV (p < 0.001 uncorrected, expected voxels per cluster k = 134).

^†Significant after FWE correction (p < 0.05, expected voxels per cluster k = 42).

Cluster peaks are specified by their anatomical site, labeled using the Hammersmith atlas provided by the CAT12 toolbox.

Brodmann areas were identified with the BioImage Suite Web 1.2.0.GMV, gray matter volume; MNI coordinates (x y z), coordinates in MNI space in millimeters; TIV, total intracranial volume.