Structural MRI scans for each patient (3-dimensional magnetization prepared rapid gradient echo, 3D MPRAGE), T1-weighted study. The left lateral views are reconstructed from the 3D MPRAGE MRI scans showing left hemisphere cortical lesion sites. Patients are listed in order of severity of aphasia from least to greatest (Table 1). Structural MRI scan for P5 was unavailable for publication. Patients P1–P7 had lesion including left Broca's cortical area and subjacent white matter. P8 had no cortical lesion. Her extensive subcortical white matter lesion included the medial subcallosal fasciculus area located deep to Broca's area, adjacent to the left frontal horn (vertical arrows); and the periventricular white matter area located deep to sensorimotor cortex mouth level, adjacent to the left body of lateral ventricle (horizontal arrow). Extensive lesion in these two white matter areas has been linked to poor prognosis of nonfluent aphasia (Naeser et al., 1989). The mildest patient, P1, had almost no lesion in these two white matter areas; the other patients had some lesion present in one, or both of these deep, white matter areas.

Schematic summary of the experimental design. Top: Picture naming was tested before and after rTMS, using Snodgrass & Vanderwart pictures (Snodgrass & Vanderwart, 1980). Bottom: Location of four right hemisphere cortical ROIs (schematically shown in lateral diagram), which were each suppressed with rTMS in separate sessions.

Effect on picture naming for right cortical ROIs that were suppressed with rTMS in the eight aphasia patients. Top (a): Change in number of pictures named versus baseline mean, post-rTMS (± standard error) to suppress each cortical ROI. Picture naming significantly increased only following rTMS to suppress the right PTr (p<.001). Bottom (b): Change in RT versus baseline mean, post-rTMS (± standard error) to suppress each cortical ROI. RT was significantly decreased after rTMS to suppress the right PTr (p<.005) and the right M1 (p<.029), but significantly increased only after rTMS to suppress the right POp (p<.052).

Effect on picture naming for left and right cortical ROIs that were suppressed with rTMS in eight normal controls. Change in RT versus baseline mean, post-rTMS (± standard error) to suppress the left PTr or POp; or right PTr or POp. Note that the RT increased maximally from baseline, after rTMS to suppress left POp, but decreased after rTMS to suppress right PTr. The RT was significantly faster (p<.009) following suppression of right PTr, versus right POp. This is consistent with the findings in the aphasia patients.