Intersubject variability of striate and extrastriate areas was mapped
by conjoined use of positron emission tomography (PET) and magnetic re
sonance imaging (MRT). We used two dynamic bowtie-shaped random-dot pa
tterns centered symmetrically around the vertical- and horizontal-meri
dian, respectively, presented during sequential PET scans in 11 subjec
ts. Control condition was simple fixation on a central dot in absence
of a surrounding random dot pattern. V1, V2, VP, V3, V3a, V4, V5, and
''wordform'' areas were identified. After spatial normalization to Tal
airach atlas space, mean locations and standard deviations about these
mean locations for x-, y-, and z-axes were calculated for each area i
n both hemispheres and compared for differences. The mean standard dev
iation for all axes across all areas tested was found to be small (4.9
mm). No significant differences were found in the mean standard devia
tions for the x-, y-, and z-axes in the left hemisphere vs. their coun
terparts in the right hemisphere. However, when mean standard deviatio
ns in both hemispheres were polled together by axis, the mean standard
deviation for the y-axis (5.3 mm) was found to be significantly diffe
rent from the mean standard deviation for the x-axis (4.3 mm). Further
more, in the left hemisphere, the mean standard deviation for the z-ax
is (5.7 mm) was significantly greater than the mean standard deviation
for the x-axis (3.9 mm). The values reported in this study for mean l
ocation and standard deviation of visual areas can be used to establis
h confidence intervals for distinguishing normal variations from patho
logy and consequent brain reorganization. Hum. Brain Mapping 6:301-315
, 1998. (C) 1998 Wiley-Liss, Inc.