Post on 26-Aug-2018
Morphology, morphometry and probability mapping of thepars opercularis of the inferior frontal gyrus: an in vivoMRI analysis
F. Tomaiuolo,1* J. D. MacDonald,2 Z. Caramanos,2 G. Posner,2 M. Chiavaras,2 A. C. Evans2 and M. Petrides2
1IRCCS `S. Lucia', Roma; Dipartimento di Scienze Neurologiche e della Visione, sez. Fisiologia Umana, Verona, Universita' diVerona, Italy2Montreal Neurological Institute,McGill University, Canada
Abstract
The pars opercularis occupies the posterior part of the inferior frontal gyrus. Electrical stimulation or damage of this region interfereswith language production. The present study investigated the morphology and morphometry of the pars opercularis in 108 normaladult human cerebral hemispheres by means of magnetic resonance imaging. The brain images were transformed into astandardized proportional steoreotaxic space (i.e. that of Talairach and Tournoux) in order to minimize interindividual brain sizevariability. There was considerable variability in the shape and location of the pars opercularis across brains and between cerebralhemispheres. There was no signi®cant difference or correlation between left and right hemisphere grey matter volumes. There wasalso no signi®cant difference between sex and side of asymmetry of the pars opercularis. A probability map of the pars operculariswas constructed by averaging its location and extent in each individual normalized brain into Talairach space to aid in localization ofactivity changes in functional neuroimaging studies.
Introduction
The posterior part of the inferior frontal gyrus in the left hemisphere
is traditionally considered to constitute the classical Broca's area, i.e.
the anterior cortical area which is thought to play a critical role in
speech production (Dejerine, 1914). The posterior half of the inferior
frontal gyrus typically exhibits two general morphological subdivi-
sions that are referred to as the pars triangularis, i.e. the triangular
portion delimited by the vertical (ascending) ramus and the horizontal
ramus of the Sylvian ®ssure, and the pars opercularis delimited by the
vertical ramus and the inferior part of the precentral sulcus (see
Fig. 1). Although it is not possible to specify more precisely the
critical zone within this general region from clinical-anatomical
correlation studies (e.g. Mohr et al., 1978), electrical stimulation
work during brain surgery has indicated that speech arrest occurs
most reliably from stimulation of that part of the inferior frontal gyrus
that lies immediately anterior to the lower end of the precentral gyrus
(Rasmussen & Milner, 1975; Ojemann, 1979).
Inspection of the caudal part of the inferior frontal gyrus in
different human brains suggests that there is considerable morpho-
logical variability. In some cases, two small convolutions can be
clearly seen on the lateral surface of the brain, lying between the
vertical ramus and the inferior precentral sulcus, whereas in other
cases only one is visible. One aim of the present study was to
examine the morphological pattern of the region of the pars
opercularis on a large-scale and to create a probability map of this
region in a standardized proportional stereotaxic space, namely that
of Talairach & Tournoux (1988) which has become the standard
space within which activation foci are expressed in most modern
functional neuroimaging studies with positron emission tomography
(PET). The provision of a probability map in which the variability in
location of a given brain region, in this case the pars opercularis, is
quanti®ed and is essential for accurate localization of the focus of
activity changes in group studies with PET. Furthermore, the
expression of the location of the average focus of activity changes
in a standardized space provides a common frame of reference across
studies.
In addition, the location of circumscribed brain lesions can be
described in a standardized space in brains that have been
appropriately transformed. Such an approach might provide a clearer
de®nition of the critical area of damage in a group of patients
exhibiting a common clinical feature (Koski et al., 1998). Thus, the
creation of a probability map of the pars opercularis will provide key
information for functional neuroanatomical studies of language.
Another aim of the present investigation was to examine whether
there are hemispheric differences in the grey matter volume of the
pars opercularis. Several studies, using different kinds of measure-
ments, examined the hemispheric and sex-related differences of the
anterior speech areas, such as the pars triangularis and pars
opercularis of the inferior frontal gyrus (e.g. Wada et al., 1975; Falzi
et al., 1982; Foundas et al., 1996, 1998). An earlier study by Wada
et al. (1975) measured the pars opercularis together with a large part
of the pars triangularis and found the overall region to be slightly
larger in the right hemisphere. However, as the authors pointed out, as
a large part of the cortex hidden within the sulci was not measured,
the possibility remained that there might still be an asymmetry in
favour of the left hemisphere if this cortex were to be included (see p.
245 of Wada et al., 1975). Witelson & Kigar (1992) reviewed the
work on the asymmetry of the frontal operculum and reported no left±
right hemispheric differences on the lateral aspect of this area, but
Correspondence: Dr Francesco Tomaiuolo, I.R.C.C.S. `Santa Lucia', ViaArdeatina 306, 00179 Rome, Italy. E-mail: f.tomaiuolo@hsantalucia.it
Received 9 December 1998, revised 19 April 1999, accepted 19 April 1999
European Journal of Neuroscience, Vol. 11, pp. 3033±3046, 1999 ã European Neuroscience Association
like Wada et al. (1975), they suggested that if the intrasulcal cortex of
the frontal operculum were to be included, an asymmetry might be
demonstrated. Falzi et al. (1982) examined the posterior portion of
the inferior frontal gyrus in 12 right-handed adults and found no
interhemispheric differences when they compared only the convexity
of this area, but a signi®cantly larger left hemisphere was found when
the comparison was performed by considering both the convexity and
the intrasulcal cortex of the posterior portion of the inferior
frontal gyrus. However, these investigators, did not include in
their measurement the pars opercularis in isolation, but together
with the pars triangularis. Recently, Foundas et al. (1996)
measured the volume of the pars triangularis on the magnetic
resonance scans of subjects who, for clinical reasons, had also
undergone an intracarotid sodium Amytal test (Wada &
FIG. 1. Sagittal (top left), coronal (top right) and horizontal (lower left) sections running through the point marked by the cross (+) in an MRI volume of a givenbrain and the reconstructed 3-D surface of the same brain (lower right). Hr, horizontal ramus of the Sylvian ®ssure; IFs, inferior frontal sulcus; Ps, precentralsulcus; Sf, Sylvian ®ssure; Vr, vertical ramus of the Sylvian ®ssure. The region of interest is marked in yellow. The arrows indicate the medialmost point of the Vrand Ps as seen on the horizontal section. The vertical and horizontal lines in purple indicate, on the 3-D surface of the brain, the level of the presented coronal andhorizontal sections, respectively.
3034 F. Tomaiuolo et al.
Ó 1999 European Neuroscience Association, European Journal of Neuroscience, 11, 3033±3046
Rasmussen, 1960) to establish the side of speech lateralization.
They found that 10 out of 11 right-handed subjects had a larger
pars triangularis and this was consistent with the side of speech
lateralization. The present investigation aimed to evaluate whether
there are hemispheric differences in the grey matter volume of the
whole pars opercularis.
FIG. 2. Cytoarchitectonic maps of the human cerebral cortex by (A) Brodmann (1909); (B) Sarkissov et al. (1955); and (C) Economo & Koskinas (1925).
Morphology and morphometry of the pars opercularis 3035
Ó 1999 European Neuroscience Association, European Journal of Neuroscience, 11, 3033±3046
Materials and methods
The MRIs of the brains of 54 randomly selected subjects were studied:
19 females aged (mean 6 SD) 25.6 6 10.1 years and 35 males aged
25.4 6 7.0 years. All the subjects were right-handed and none had a
positive history for neurological or psychiatric disorder. The subjects
gave informed consent. The MRI scans were made by means of a
Philips Gyroscan 1.5-T superconducting magnet system using a 3-D
fast-®eld echo-acquisition sequence T1-weighted (TR, 18 ms, TE,
10 ms and ¯ip angle 30 °), collecting 160 slices of 1 mm thickness, in
the sagittal plane.
In order to standardize the images and correct for any interindividual
differences in gross brain size, each MRI volume was separately
transformed into the standardized proportional stereotaxic space of
Talairach & Tournoux (1988) by means of an automatic registration
program which uses a 3-D cross-correlation approach to match the
3036 F. Tomaiuolo et al.
Ó 1999 European Neuroscience Association, European Journal of Neuroscience, 11, 3033±3046
FIG. 3. (a and b) The pars opercularis (identi®ed by the yellow colour), as visualized from the lateral view of the 3-D reconstructed left hemisphere of each one ofthe brains that were studied. The hemispheres are presented on a grid where the stereotaxic Z-values range from +50 mm (ventral) to +70 mm (dorsal) and the Y-values range from ±30 mm (caudal) to +50 mm (rostral). The surface rendering of the left hemisphere is shown at ±30 mm lateral to the midline. The sulci thatde®ne the pars opercularis can be identi®ed on these grids as follows. The sulcus just caudal to the region indicated in yellow is the inferior precentral sulcus, theone dorsal is the inferior frontal sulcus, the one ventral is the Sylvian ®ssure and the one just rostral is the vertical ramus of the Sylvian ®ssure. The sulcusdiagonalis, when it is present, can be seen within the yellow region.
Morphology and morphometry of the pars opercularis 3037
Ó 1999 European Neuroscience Association, European Journal of Neuroscience, 11, 3033±3046
single MRI volume with the intensity average of 305 MRI brain
volumes previously aligned into the standardized stereotaxic space
(Collins et al., 1994). This resampling results in a volume of 160 axial
slices with an in-plane matrix of 256 3 256 pixels. The thickness of the
interpolated sagittal, axial and coronal slices were 0.67, 0.75 and
0.86 mm, respectively. It is important to note that transformation into
the Talairach & Tournoux (1988) stereotaxic space is the usual
procedure for averaging functional activity obtained in different brains
with PET in order to describe the location of the activation peaks.
Region of interest: anatomical landmarks and boundaries
The pars opercularis can be delimited macroscopically as follows:
rostrally, by the vertical (ascending) ramus of the Sylvian ®ssure
(Vr); dorsally, by the inferior frontal sulcus (IFs); caudally, by the
3038 F. Tomaiuolo et al.
Ó 1999 European Neuroscience Association, European Journal of Neuroscience, 11, 3033±3046
inferior part of the precentral sulcus (Ps); ventrally, by the Sylvian
®ssure (Sf). Cytoarchitectonic analyses have indicated that the
surface area of the pars opercularis is characterized by a distinct
type of cortex that was labelled as area 44 by Brodmann (1908) and
Sarkissov et al. (1955) and area FCBm by Economo & Koskinas
(1925) (see Fig. 2).
FIG. 4. (a and b) The pars opercularis (identi®ed by the yellow colour), as visualized from the lateral view of the 3-D reconstructed right hemisphere of each one ofthe brains that were studied. The hemispheres are presented on a grid where the stereotaxic Z-values range from +50 mm (ventral) to +70 mm (dorsal) and the Y-values range from ±30 mm (caudal) to +50 mm (rostral). The surface rendering of the right hemisphere is shown at +30 mm lateral to the midline. The sulci thatde®ne the pars opercularis can be identi®ed on these grids as follows. The sulcus just caudal to the region indicated in yellow is the inferior precentral sulcus, theone dorsal is the inferior frontal sulcus, the one ventral is the Sylvian ®ssure and the one just rostral is the vertical ramus of the Sylvian ®ssure. The sulcusdiagonalis, when it is present, can be seen within the yellow region.
Morphology and morphometry of the pars opercularis 3039
Ó 1999 European Neuroscience Association, European Journal of Neuroscience, 11, 3033±3046
This area can be distinguished from the posteriorly adjacent
premotor cortex (Brodmann's ventral area 6 or Economo and
Koskinas' area FB), which lies on the precentral gyrus, by the
presence of an incipient layer IV and very large pyramidal neurons in
the deeper portion of layer III. Rostrally, area 44 is replaced by
Brodmann's area 45 (Economo and Koskinas' area FDG) which has a
well-developed layer IV. The cytoarchitectonic studies of Petrides &
Pandya (1994) have shown that the surface of the cortex occupying
the pars opercularis is area 44 in all brains and that this area extends
into the adjacent sulci (i.e. Vr, IFs and Ps) for a considerable distance,
although the precise limit varies across brains. However, the cortex
within the Sylvian ®ssure that joins the pars opercularis with the
insula is a different architectonic area (see Fig. 2C; Economo and
Koskinas, 1925) and was therefore not included in the region of
interest. Thus, as can be seen in Fig. 1, the region of interest in this
study included, apart from the cortex on the surface of the brain, the
cortex of the posterior bank of the vertical ramus of the Sylvian
®ssure, the cortex on the lower bank of the inferior frontal sulcus and
the cortex of the anterior bank of precentral sulcus up to the level of
the inferior frontal sulcus. We de®ned the ventral limit of the region
of interest as a straight line running from the ventromedial end of the
vertical ramus up to the ventralmost part of the precentral sulcus.
When the diagonal sulcus was present at that position, the cortex
lining it was also included as part of the region of interest. A straight
line was drawn from the dorsalmost point of the vertical ramus up to
the inferior frontal sulcus to delimit the dorsorostral limit of the
region of interest, as the vertical ramus does not reach the inferior
frontal sulcus. Similarly, whenever the inferior frontal sulcus did not
merge with the precentral sulcus, a straight line was drawn from the
dorsalmost point of the inferior frontal sulcus to the precentral sulcus
to delimit the dorsocaudal limit of the region of interest.
Grey-matter labelling in the region of interest
The grey matter in the region of interest was mapped using
DISPLAY, an interactive program developed at the Montreal
Neurological Institute by J. D. MacDonald. This program permits
labelling of voxel regions on each slice of the MRI volume and
allows for the simultaneous visualization of the movement of the
cursor on the screen within the sagittal, horizontal and coronal
planes of the MRI. The movement of the cursor can also be
viewed on a 3-D reconstruction of the brain surface (see Fig. 1).
The reconstruction of the surface in 3-D of each brain was made
by means of a 3-D model-based surface deformation algorithm
(MacDonald, 1998).
The grey matter was estimated by selecting a range for the
contrast intensity voxels following a DISPLAY procedure that
produces a histogram of the full-brain MRI volume contrast
intensity. The plotting of the contrast intensity for the full brain
results, typically, in three peaks; the ®rst indicates the `cerebro-
spinal' ¯uid, the second the `grey matter' and the third the `white
matter'. The grey-matter range was determined by identifying the
midpoint between the ®rst and the second peaks and the midpoint
between the second and the third peaks. The MRI intensities are
imperfectly correlated with tissue type and always result in
relative, not absolute, values (see Clarke et al., 1995; for review).
These intensity values, however, should affect grey and white
matter de®nition equally, and should therefore be consistent across
the hemispheres of the same brain (for discussion see Paus et al.,
1996 and Penhune et al., 1996). A selection of any area can be
performed by using the DISPLAY `mouse-brush' that marks the
voxels, by colouring. This colouring procedure accompanied by
the simultaneous 3-D view of the MRI planes allows a better
identi®cation of the cortex contained in the depths of the sulci
(see Fig. 1). The region of interest was ®rst identi®ed by the
landmarks outlined above; then the area to be marked was
brushed by the `mouse brush' in DISPLAY. This act resulted in
the selection by DISPLAY of those voxels previously de®ned as
grey matter according to the established contrast ranges.
Generation of the probability map
The probability map of the region of interest was generated by
averaging the labelled voxels of the individual grey matter volumes
across the different normalized brains in the Talairach space.
Statistical analyses
We carried out separate statistical analyses of the volumes obtained in
both the transformed standardized space (i.e. Talairach & Tournoux,
1988) and the nontransformed native space. The latter represents the
original brain image obtained from each subject. In this manner, it
was possible to obtain a better between-subject evaluation. By using
the standardized space, size and site variability between subjects is
reduced. On the other hand, the use of the native space permits a
better comparison with previous investigations of the morphometry of
the inferior frontal gyrus (e.g. Wada et al., 1975; Falzi et al., 1982).
To produce the voxel volumes in native space for each hemisphere,
the procedure used to transform the brains in standardized space was
inverted.
A correlation between the region-of-interest subject volumes
obtained in standardized and in native spaces was carried out. In
addition, the correlations between the left and right hemisphere on the
region of interest volumes obtained in standardized and in native
space were carried out. Interhemispheric grey-matter asymmetries
were calculated on the grey matter volumes obtained in standardized
and in native space separately with the asymmetry index used by
Rademacher et al. (1993). The index was computed for each subject
as follows: asymmetry = (left volume ± right volume)/[(left volume +
right volume)/2]. Then, a 2 3 2 c2 test (sex and side of the asymmetry
index) was carried out for each space.
Moreover, two-factor ANOVA with the between-subject factor being
sex (female vs. male) and the within subject factor being side (left vs.
right hemisphere) in region of interest volumes, both in standardized
and in native space, were carried out.
Results
The vertical ramus could not be identi®ed in two of the cerebral
hemispheres studied; this was also the case with the precentral sulcus
in two different hemispheres. These four brains were excluded from
the ®nal analysis and, therefore, the results reported below are based
on 100 cerebral hemispheres studied in 50 subjects. The region of
interest was labelled bilaterally in the brains of 17 females aged
24.97 6 10.19 years and 33 males aged 25.63 6 7.18 years. There was
no signi®cant difference in age between males and females
(t(48) = 0.266).
Morphological ®ndings
The morphology of the pars opercularis, as visualized from the
lateral view of each one of the 100 hemispheres examined, is
presented in Fig. 3 (a and b) for the left hemisphere and in Fig. 4
(a and b) for the right hemisphere. The bright yellow represents
the part which can be observed on the surface of the brain,
whereas the pale yellow represents the part which is hidden
within the sulci and can therefore not be visualized from the
surface of the brain. As can be seen in Figs 3 and 4, there was
3040 F. Tomaiuolo et al.
Ó 1999 European Neuroscience Association, European Journal of Neuroscience, 11, 3033±3046
considerable variability in the shape of the pars opercularis. In
some of the brains, almost the entire pars opercularis, appearing
as a single or a double gyrus, was visible from the surface of the
brain (Figs 1 and Fig. 5a); in other brains, a variable part of the
pars opercularis was submerged in the inferior precentral sulcus
(see Fig. 5b) and, in four left hemispheres and one right
hemisphere, there was only one gyrus interspersed between the
vertical ramus and the precentral sulcus, and this gyrus was
completely submerged (see Fig. 3b, right half, subjects 34, 39, 41
and 42, and Fig. 4b, right half, subject 49, and also Fig. 6a and b).
In such cases, the vertical ramus and the precentral sulcus could
not be distinguished on the lateral surface of the brain (see red
arrow in Fig. 6a and b), but they could be distinguished on axial
sections of the brain (see white arrows in Fig. 6a and b).
FIG. 5. (a) An example of a brain in which the entire pars opercularis consisted of a double gyrus. (b) An example of a brain with a variable part of the parsopercularis submerged in the precentral sulcus. The solid blue line indicates the level of the horizontal section in b.
Morphology and morphometry of the pars opercularis 3041
Ó 1999 European Neuroscience Association, European Journal of Neuroscience, 11, 3033±3046
FIG. 6. (a and b) Two examples of brains in which there is only one gyrus interspersed between the vertical ramus (Vr) and the inferior precentral sulcus (Ps), andthis gyrus was completely submerged. Note that in such cases, the vertical ramus and the inferior precentral sulcus cannot be distinguished on the lateral surface ofthe brain (see red arrow), but they can easily be identi®ed on the horizontal sections of the brain (see white arrows).
3042 F. Tomaiuolo et al.
Ó 1999 European Neuroscience Association, European Journal of Neuroscience, 11, 3033±3046
Morphometric ®ndings
Correlations
Across the different brains, there was an almost perfect linear
relationship between the grey-matter volumes of the region of interest
in standardized stereotaxic space and those in native space. This was
true both for the region-of-interest volumes labelled in the left
hemisphere (r2 = 0.89, P < 0.001) and for those labelled in the right
hemisphere (r2 = 0.91, P < 0.001). This merely re¯ects the fact that
values in standardized and in native space are simply transformations
of one another.
As can be seen in Fig. 7, there was no signi®cant linear relationship
in the grey-matter volumes in the region of interest between the left
and the right hemispheres. This was true in both standardized space
(Fig. 7: r2 = 0.07, P = 0.059) and in native space (r2 = 0.02,
P = 0.318).
An asymmetry score was calculated for each brain. A negative
asymmetry score indicates that the volume of the left hemisphere is
smaller than that of the right hemisphere, whereas a positive score
indicates a smaller right-side volume. The distribution of interhemi-
spheric asymmetry scores, grouped by sex, as well as the group
means and standard deviations are shown in Fig. 8. There was no
effect of sex on the distribution of which hemisphere had a greater
overall grey-matter volume (c2 = 0.677, P = 0.41).
Two separate ANOVAs for the volumes in standardized and in native
space, respectively, did not reveal signi®cant differences between the
male and the female brains (F1,48 = 1.22, for standardized space;
F1,48 = 0.066, for native space) in region-of-interest volumes or in the
crucial comparison between left and right hemispheres (F1,48 = 1.222,
for standardized space; F1,48 = 0.868, for native space). No signi®cant
interactions were obtained in either analyses (F1,48 = 0.114, for
standardized space; F1,48 = 0.065, for native space; see Fig. 9a and b).
Probability map
The probability map of the pars opercularis constructed from the
present analysis is shown in Fig. 10. This ®gure presents contour
outlines for 18 coronal sections overlaid on grids corresponding to the
Talairach & Tournoux (1988) stereotaxic space. These grids are taken
2 mm apart in the rostrocaudal direction (Y). Y represents the distance
of the coronal section rostral from the anterior commissure which, in
the Talairach & Tournoux (1988) space, is de®ned as Y = 0. The
dorsoventral direction (Z) is shown on the ordinate of the grid. The
positive numbers represent distance in mm dorsal to the anterior
commissure-posterior commissure horizontal line, which is de®ned as
Z = 0, and the negative numbers refer to distance ventral to this line.
The medial-to-lateral (X) distance relative to the midline (X = 0) is
shown on the abscissa (positive numbers indicate right hemisphere
and negative numbers left hemisphere). The white, grey, and black
areas in the contour outline represent different ranges of frequency of
occurrence of a given voxel in the region of interest (white = 5%-
25%, grey = 25%-50%, black = 50%-75%). None of the voxels had a
frequency of occurrence higher than 74%.
Discussion
Variability of the region of the pars opercularis
The present investigation examined the morphology of the pars
opercularis of the inferior frontal gyrus in a large sample of normal
human adult brains. This region, which is de®ned as the part of the
inferior frontal gyrus that lies between the inferior precentral sulcus
and the vertical ramus of the Sylvian ®ssure, exhibits considerable
morphological variability. As can be seen in Figs 3 and 4, in some of
the brains, the pars opercularis consists of two small convolutions,
whereas in other brains only one can be identi®ed. Furthermore, in
some of the cases where there are two convolutions, one of these may
lie on the surface of the brain while the other may lie hidden within
the depth of the inferior precentral sulcus (see Fig. 5). This
convolution may be completely or partially submerged. In four left
hemispheres and one right hemisphere, there was only one
convolution interspersed between the vertical ramus and the inferior
precentral sulcus, and this was completely submerged. In such cases,
the vertical ramus and the inferior precentral sulcus could not be
distinguished on the lateral surface of the brain (see red arrow in
FIG. 7. Graphs illustrating the lack of a signi®cant linear relationship betweenthe left and right hemisphere grey-matter volumes in the region of interest inthe subjects studied. The data presented are in standardized space.
FIG. 8. The distribution of interhemispheric asymmetry scores, grouped bysex, and the group means and standard deviations. A negative asymmetryscore indicates a smaller left-sided volume; a positive score indicates a smallerright-sided volume.
Morphology and morphometry of the pars opercularis 3043
Ó 1999 European Neuroscience Association, European Journal of Neuroscience, 11, 3033±3046
Fig. 6), but they could easily be identi®ed on horizontal sections of
the brain (see white arrows in Fig. 6).
These ®ndings have relevance to the identi®cation of the anterior
speech area by means of electrical stimulation in patients undergoing
brain surgery. Rasmussen & Milner (1975) pointed out that speech
arrest or interference with speech can occur as a result of electrical
stimulation `from one or the other, or both of the two frontal
opercular convolutions immediately anterior to the lower end of the
precentral gyrus' (see p. 241 of their paper). Ojemann (1979) also
emphasized that this region of the inferior frontal gyrus is the part
from which speech arrest with brain stimulation can most reliably be
elicited, but he went on to stress the variability of the precise location
within this region from which speech arrest can occur. Although
some of this variability may be due to variability in the stimulation
parameters and the nature of the behavioural response required, the
possibility that this variability also re¯ects morphological differences
must seriously be entertained. For instance, in brains where all or a
part of the pars opercularis is submerged in the inferior precentral
sulcus, electrical stimulation may evoke motor responses as a result
of stimulating the precentral gyrus, just posterior, but no or only
minor speech interference may be evoked from stimulation
immediately in front of it because the critical area may be lying
largely within the precentral sulcus.
Diagonal sulcus
The diagonal sulcus is a small sulcus that is present in the pars
opercularis (Eberstaller,1890; Bailey & Bonin, 1951). It may lie
within the pars opercularis without joining any of the sulci that
surround the pars opercularis, or it may join, dorsally, the inferior
frontal sulcus or the inferior precentral sulcus, caudally. The
lower end of the diagonal sulcus may ¯ow into the Sylvian
®ssure or it may stay clearly separate from it, separated by a
bridge of cortex (Eberstaller, 1890; in Bailey & Bonin, 1951).
Whereas Turner (1948) mentioned that the diagonal sulcus is a
regular feature of the human brain, Bailey & Bonin (1951)
reported that it is by no means a constant feature of the human
brain. Recently Ono et al. (1990) reported the presence of the
diagonal sulcus in 64% of the right and 72% of the left cerebral
hemispheres that they examined. Galaburda (1980), however, on
the basis of inspection of the photographs of 102 brains studied
by Geschwind & Levitsky (1968) found the diagonal sulcus in
26.5% in the left hemisphere and 12.75% in the right one. In the
present study, the diagonal sulcus, as shown in the maps of
Brodmann (1908, 1909) and Economo & Koskinas (1925) could
clearly be identi®ed in 34% of the left and 32% of the right
hemispheres. In addition, we observed a small dimple between the
inferior precentral sulcus and the vertical ramus in almost all the
hemispheres studied. The discrepancies in the reported incidence
of the diagonal sulcus are probably due to the different criteria in
de®ning it.
Morphometry
Although Wada et al. (1975) demonstrated a rightward asymmetry
in the region of the inferior frontal gyrus that included the pars
opercularis and a part of the pars triangularis, they suggested that
a leftward asymmetry in this region might have been observed if
the cortex lying in the sulci were to be included in the
measurements. Indeed, Falzi et al. (1982) reported a leftward
FIG. 9. Distribution of grey matter volumes of the pars opercularis across sex (females±males) and hemisphere (left±right). The circles represent the volumes foreach hemisphere in standardized stereotaxic space (a) and in nontransformed native space (b).
3044 F. Tomaiuolo et al.
Ó 1999 European Neuroscience Association, European Journal of Neuroscience, 11, 3033±3046
asymmetry in this region, as did Foundas et al. (1996, 1998) in
separate measurements of the pars triangularis and pars oper-
cularis. The present investigation, in which measurements were
made both of cortex lying on the lateral surface of the pars
opercularis and cortex hidden in the sulci, did not reveal
differences in cortical volume between the left and right
hemispheres. The above discrepancies may have been due to
the different extent to which speci®c cytoarchitectonic areas may
have been included in the measurements. For instance, Galaburda
(1980) reported that six out of 10 brains had a leftward
asymmetry in area 44.
Probability map
The present study has provided a probability map of the pars
opercularis within the Talairach and Tournoux proportional stereo-
taxic space (see Fig. 10). This map can be used to localize activation
foci obtained in functional neuroimaging studies, within the pars
opercularis region of the inferior frontal gyrus. The classical model of
language organization in the human brain, which is based on clinical±
anatomical correlations, posits the existence of an anterior speech
area, located on the posterior part of the inferior frontal gyrus in the
dominant hemisphere, that is involved in speech production (see
FIG. 10. Probability contour maps of grey-matter of pars opercularis in a series of coronal sections displayed on a grid (Y, X, Z). White, grey, and black bordersindicate the occurrence frequencies of 5±25%, 25±50% and 50±75%, respectively.
Morphology and morphometry of the pars opercularis 3045
Ó 1999 European Neuroscience Association, European Journal of Neuroscience, 11, 3033±3046
Geschwind, 1970). Electrical stimulation studies during brain surgery
indicate that the critical region for evoking speech interference is in
the one or two convolutions immediately anterior to the precentral
sulcus (Rasmussen & Milner, 1975; Ojemann, 1979). However,
functional activation studies with PET have demonstrated that
activation related to speech varies considerably between different
studies, suggesting that the activation focus or foci within the inferior
frontal lobe of the dominant hemisphere may vary depending on the
speci®c requirement of the language task used (e.g. Petersen et al.,
1988, 1990; Paulesu et al., 1993, 1997; Petrides et al., 1993). The
availability of a probability map therefore becomes critical for
establishing, with a known probability, whether the activation focus
lies within the pars opercularis or whether it lies just anterior, in the
pars triangularis, or just posterior, in the premotor cortex of the lower
precentral gyrus.
Abbreviations
Hr, horizontal ramus of the Sylvian ®ssure; IFs, inferior frontal sulcus; PET,positron emission tomography; Ps, precentral sulcus; Sf, Sylvian ®ssure; Vr,vertical ramus of the Sylvian ®ssure.
References
Bailey, P. & Bonin, G. (1951) The Isocortex of Man. University of IllinoisPress, Urbana.
Brodmann, K. (1908) Beitraege zur histologischen Lokalisation derGrosshirnrinde. VI. Mitteilung: die Cortexgliederung Des Menschen. J.Psychol. Neurol. (Lzp.), 10, 231±, 246.
Brodmann, K. (1909) Vergleichende Lokalisationslehre der Grosshirnrinde inihren Prinzipien dargestellt auf Grund des Zellenbaues. Barth, Leipzig.
Clarke, L.P., Velthuizen, R.P., Camacho, M.A., Heine, J.J., Vaidyanathan, M.,Hall, L.O., Thatcher, R.W. & Silbiger, R.W. (1995) MRI segmentation:Methods and applications. Magn. Reson. Imag., 13, 343±368.
Collins, D.L., Neelin, P., Peters, T.M. & Evans, A.C. (1994) Automatic 3Dintersubject registration of MR volumetric data in standardized Talairachspace. J. Comput. Assist. Tomogr., 18, 192±205.
Dejerine, J.J. (1914) Semiologie Des Affections Du Systeme Nerveux. Massonet Cie, Paris.
Eberstaller, O. (1890) Das Stirnhirn. Urban & Schwarzenberg, Wien.Economo, C. & Koskinas, G.N. (1925) Die Cytoarchitektonik der Hirnrinde
Des Erwachsenen Menschen. Springer, Wien.Falzi, G., Perrone, P. & Mignolo, L.A. (1982) Right±left asymmetry in
anterior speech region. Arch. Neurol., 39, 239±240.Foundas, A.L., Eure, K.F., Luevano, L.F. & Weinberger, D.R. (1998) MRI
asymmetries of Broca's area: The pars triangularis and pars opercularis.Brain Lang., 64, 282±296.
Foundas, A.L., Leonard, C.M., Gilmore, R.L. & Fennel, E.B. (1996) Parstriangularis asymmetry and language dominance. Proc. Natl Acad. Sci.USA, 93, 719±722.
Galaburda, A.M. (1980) La reÂgion de Broca: Observations anatomiques faitesun sieÁcle apreÁs la mort de son deÂcouvreur. Rev. Neurol., 136, 609±616.
Geschwind, N. (1970) The organization of the language and the brain. Science,170, 940±944.
Geschwind, N. & Levitsky, W. (1968) Human brain: Left-right asymmetries intemporal speech region. Science, 161, 186±187.
Koski, L.M., Paus, T. & Petrides, M. (1998) Directed attention after unilateralfrontal excisions in humans. Neuropsychologia, 36, 1363±1371.
MacDonald, J.D. (1998) A method for identifying geometrically simplesurfaces from three dimensional images. PhD Dissertation, McGillUniversity, Montreal, Canada.
Mohr, J.P., Pessin, M.S., Finkelstein, S., Funkenstein, H.H., Duncan, G.W. &Davis, K.R. (1978) Broca aphasia: pathologic and clinical. Neurology, 28,311±324.
Ojemann, G.A. (1979) Individual variability in cortical localization oflanguage. J. Neurosurg., 50, 164±169.
Ono, M., Kubik, S. & Adernathey, C.D. (1990) Atlas of the Cerebral Sulci.Thieme, New York.
Paulesu, E., Frith, C.D. & Frackowiak, R.S. (1993) The neural correlates of theverbal component of working memory. Nature, 362, 342±345.
Paulesu, E., Goldacre, B., Scifo, P., Cappa, S.F., Gilardi, M.C., Castiglioni, I.,Perani, D. & Fazio, F. (1997) Functional heterogeneity of left inferiorfrontal cortex as revealed by fMRI. Neuroreport, 8, 2011±2017.
Paus, T., Otaky, N., Caramanos, Z., MacDonald, J.D., Zijdenbos, A.,D'Arrivo, D., Gutmans, D., Holmes, C., Tomaiuolo, F. & Evans, A.C.(1996) In-vivo morphometry of the intrasulcal gray matter in the humancingulate, paracingulate, and superior-rostral sulci: hemisphericasymmetries, gender differences and probability maps. J. Comp. Neurol.,376, 664±673.
Penhune, V.B., Zatorre, R., MacDonald, J.D. & Evans, A.C. (1996)Interhemispheric anatomical differences in human primary auditorycortex: Probability map and volume measurement from magneticresonance scans. Cereb. Cortex, 6, 661±672.
Petersen, S.E., Fox, P.T., Posner, M.I., Mintun, M. & Raichle, M.E. (1988)Positron emission tomographic studies of the cortical anatomy of single-word processing. Nature, 331, 585±589.
Petersen, S.E., Fox, P.T., Snyder, A.Z. & Raichle, M.E. (1990) Activation ofextrastriate and frontal cortical areas by visual words and word-like stimuli.Science, 249, 1041±1044.
Petrides, M., Alivisatos, B., Meyer, E. & Evans, A.C. (1993) Functionalactivation of the human frontal cortex during the performance of verbalworking memory tasks. Proc. Natl Acad. Sci. USA, 90, 878±882.
Petrides, M. & Pandya, D.N. (1994) Comparative architectonic analysis of thehuman and the macaque frontal cortex. In Boller, F. & Grafman, J. (eds),Handbook of Neuropsychology, Vol. 9. Elsevier, Amsterdam, pp. 17±58.
Rademacher, J., Caviness, V.S., Seteinmetz, H. & Galaburda, A.M. (1993)Topographical variation of the human primary cortices: complications forneuroimaging. Brain mapping and neurobiology. Cereb. Cortex, 3, 313±329.
Rasmussen, T. & Milner, B. (1975) Clinical and surgical studies of thecerebral speech areas in man. In Zulch, K.J.Creutzfeldt, O. & Galbraith,G.C. (eds), Cerebral Localization. Springer-Verlag, Berlin, pp. 238±257.
Sarkissov, S.A., Filimonoff, I.N., Kononowa, E.P., Preobraschenskaja, I.S. &Kukuew, L.A. (1955) Atlas of the Cytoarchitectonics of the HumanCerebral Cortex. Medgiz, Moscow.
Talairach, J. & Tournoux, P. (1988) Co-Planar Stereotactic Atlas of theHuman Brain. Gerorg Thieme-Verlag, Stuttgart.
Turner, O.A. (1948) Growth and development of the cerebral cortical patternin man. Arch. Neurol. Psych., 59, 1±12.
Wada, J. & &. Rasmussen, T. (1960) Intracarotid sodium amytal for thelateralization of cerebral speech dominance. J. Neurosurg., 17, 266±282.
Wada, J., Clarke, R. & Hamm, A. (1975) Cerebral henispheric asymmetry inhumans. Arch. Neurol., 32, 239±246.
Witelson, S.F. & Kigar, D.L. (1992) Sylvian ®ssure morphology andasymmetry in man and woman: Bilateral differences in relation tohandedness in man. J. Comp. Neurol., 323, 326±340.
3046 F. Tomaiuolo et al.
Ó 1999 European Neuroscience Association, European Journal of Neuroscience, 11, 3033±3046