Normal Variation in the Frequency and
Location of Human Auditory Cortex
Landmarks. Heschl’s Gyrus: Where Is It?
Christiana M. Leonard1,2, Cynthia Puranik3, John M. Kuldau2
and Linda J. Lombardino3
Interpersonal communication via the auditory modality is fundamental to normal human development. One of the prominent
anatomical specializations supporting this communication is the
transverse gyrus of Heschl on the superior surface of the temporal
lobe. This gyrus frequently appears duplicated, either by a sulcus
indenting the crown of an initially single gyrus (common stem), or
by a complete posterior duplication. The frequency of these
duplications has been reported to be elevated in populations with
learning disabilities and genetic anomalies. The significance of this
observation is unclear, however, due to conflicting reports of the
base rate of duplication and the location of relevant sulcal
landmarks. In this study we report the variation in frequency and
location of the sulcal boundaries of Heschl’s gyrus in volumetric
magnetic resonance imaging scans of 105 normal controls aged
5–65. The major results were as follows: (i) duplications were
unstable – the frequency of duplication ranged from 20 to 60%
depending on distance from the midline; (ii) common stem duplications were more frequent than posterior duplications, particularly
in the right hemisphere. Intra- and interindividual instability in sulcal
landmarks pose serious obstacles to the attempt to map behavioral
function onto the brain. Novel methods for dealing with structural
variation are needed to facilitate the development of valid mapping
techniques.
and H2S on the right as anterior boundaries, due to their belief
that duplications were standard on the right. Subsequent authors
have tended to follow Witelson and chosen H1S on both sides
(Witelson, 1982). Steinmetz et al. (1989) were the first modern
authors to call attention to SI, although they did not consider it to
be the anterior boundary of the planum. In the most recent
study, Penhune et al. (1996) accepted SI as an anterior border if
it indented more than half the lateral extent of H. The frequency
of duplications reported in the literature ranges from 15 to 83%
on the right and from 10 to 40% on the left (Von Economo and
Horn, 1930; Campain and Minckler, 1976; Musiek and Reeves,
1990; Penhune et al., 1996). Inconsistency in reported frequencies could ref lect real differences between the populations
sampled or be an artifactual result of technical differences in the
definition of sulcal landmarks.
A lthough it is conventional to refer to the first transverse
gyrus of Heschl as the site of primary auditory cortex, recording
experiments in a small number of humans undergoing surgery
for epilepsy have revealed that primary auditory cortex
(as defined by response latency) is actually restricted to the
caudomedial region of the gyrus (Liegeois-Chauvel et al., 1991).
Thus measurements of Heschl’s gyrus which include its entire
length overestimate the extent of primary auditor y cortex.
Nonprimary areas of H1G, H2G and the remaining superior
temporal gyrus are thought to contain duplicated auditory maps
specialized in the processing of particular auditory features
(Merzenich and Brugge, 1973; Rauschecker et al., 1995),
analogous to those found in the visual system (Allman and Kaas,
1971; Allman, 1987; Felleman and Van Essen, 1991; Sereno et
al., 1995). In animals as diverse as canaries and bats (Hauser,
1996), it is known that distinctive features of critically important
species-specific signals are amplified in topographically
organized maps. By analogy, it seems reasonable to suppose that
the size and diversity of auditory maps on Heschl’s gyrus and the
planum could be related to the efficiency of auditory
and language processing in the human. For this reason, it is
important to obtain information on the sampling distribution of
auditor y cortex landmarks, as variation in these anatomical
characteristics might serve as a visible indicator of individual
variation in physiology and behavior.
In postmortem studies, it is rarely possible to relate variation
in anatomy to behavioral characteristics during life [although
Witelson and Kigar’s (1992) prospective study of terminal
cancer patients is a notable exception]. The introduction of
high-resolution magnetic resonance technology has allowed the
establishment of a new field: in vivo human cognitive
neuroanatomy — the use of magnetic resonance imaging (MRI)
scans for the identification of the anatomical substrate of human
behavioral variation (Bookstein, 1996). A number of studies have
focused on Heschl’s gyrus and the planum temporale, due to the
pioneering work of Galaburda (1989). Studies using qualitative
Introduction
One of the most prominent landmarks in the human brain is the
bulge in the lower bank of the sylvian fissure that is formed by
auditory cortex on the transverse gyrus of Heschl (H, see Fig. 1).
Posterior to this bulge is a prominent sulcus [Heschl’s sulcus
(H1S)] and the planum temporale (P), the only structure that is
reliably larger in the left hemisphere than the right (Geschwind
and Levitsky, 1968; Witelson, 1982; Steinmetz et al., 1990b;
Leonard et al., 1993; Kulynych et al., 1994). Agreement on the
direction of planar asymmetry is remarkable since both the
anterior and the posterior boundary of the planum can be
difficult to identify, due to anatomical variation. The location of
the anterior boundary is affected by duplication of H. Two types
of duplication have been described: in one case, a sulcus
intermedius (SI) indents the crown, forming a common stem
(CS) duplication (see Fig. 2); in the other, depicted in Figure 3,
there is a complete posterior duplication (PD) of H that is
bounded by a second Heschl’s sulcus (H2S) (Rademacher et al.,
1993).
A recent cytoarchitectonic study of a small number of brains
(Rademacher et al., 1993) has reported that the granular
koniocortex characteristic of primary cortex is limited to the
first H in both types of duplication. This suggests that if the
planum temporale is to be defined as the cortical zone caudal to
primary auditory cortex, its anterior boundary should always be
the most anterior sulcus associated with H, i.e. the SI if it exists,
or H1S, in the case of a PD. This rule has rarely been followed in
the past. Geschwind and Levitsky (1968) chose H1S on the left
Departments of Neuroscience, 2Psychiatry and
Communication Sciences, University of Florida, Gainesville,
FL 32610, USA
1
3
Cerebral Cortex Jul/Aug 1998;8:397–406; 1047–3211/98/$4.00
Table 1
Demographic characteristics of sample stratified by age
Children
Age (years)
Handedness index
Nonverbal intelligencea
Phonemic awarenessb
Adults
n
Mean ± SD
n
Mean ± SD
52
52
40
50
8.9 ± 2.0
0.64 ± 0.57
107.2 ± 13.1
69.8 ± 22.6
53
53
30
50
33.8 ± 11.1
0.29 ± 0.74
108.3 ± 13.3
90.5 ± 14.3
a
The children were given the Test of Nonverbal Intelligence (Brown et al., 1982), the adults were
given the Wechsler Scale of Adult Intelligence. The mean performance IQ score is given here.
b
Phonemic awareness was measured with the Lindamood Test of Auditory Conceptualization
(Lindamood and Lindamood, 1979).
1995). A quantitative study found that young children with poor
phonemic awareness had significantly less planar asymmetry
than their same-aged peers with good phonemic skills (Leonard
et al., 1996). Reduced planar asymmetry has also been reported
in women and left-handers (Witelson and Kigar, 1992; Steinmetz
et al., 1991; Foundas et al., 1995) and in children with language
impairment (Gauger et al., 1997), while exaggerated planar
asymmetry is associated with absolute pitch and early musical
ability (Schlaug et al., 1995).
In view of the conf licting data on the frequency of duplication
and its importance for the calculation of planar asymmetry, we
decided to investigate the incidence of duplication in a large set
of normal controls. Our objectives were to determine (i) the
anterior and posterior borders of Heschl’s gyrus and the planum
temporale in Talairach space; (ii) the frequency of the two types
of duplication of Heschl’s gyrus in the left and right hemispheres; and (iv) the contribution of age, sex and handedness to
variation in the location of these sulcal landmarks.
Materials and Methods
Subjects
The data were acquired from archived scans of individuals who participated as controls in funded studies. The scans of 52 children aged 5–12
(29 boys, 23 girls) and 53 adults aged 18–65 (40 men, 13 women) were
used. All subjects had signed informed consent forms after the purpose
and risks of the studies were explained.
Figure 1. A single Heschl’s gyrus at x = 46 in sagittal (a,c) and coronal (b) planes of
section of the left hemisphere. White mark locates intersection point of the two planes.
(a) A long planum is typical in the left hemisphere. (b) Planum forms superior bank of the
superior temporal gyrus. (c) Magnification of (a). Arrowheads indicate sulcal landmarks
located in all brains. Following Geschwind and Levitsky (1968), most studies have
measured the surface of the planum between H1S and PHRP. Abbreviations for all
figures: Ax, axial plane of section; H, the (first) transverse gyrus of Heschl; HA, the
sulcus anterior the first transverse gyrus of Heschl; H1S, the sulcus posterior to the
(first) transverse gyrus of Heschl; P, planum temporale; PHRP, the termination of the
posterior horizontal ramus; STG, superior temporal gyrus; Thal, thalamus; Sag, sagittal;
Cor. coronal.
categorical methods adopted from Steinmetz et al. (1990a) and
Witelson and Kigar (1992) have reported that duplicated
Heschl’s gyri were increased in families with learning disabilities
(Leonard et al., 1993; Mercado et al., 1994) as well as in boys
with a genetic resistance to thyroid hormone (Leonard et al.,
398 Location of Auditory Cortex Landmarks • Leonard et al.
Test Battery
Subjects received a handedness preference questionnaire (Briggs et al.,
1976) from which a quantitative handedness score was calculated (–1 =
completely left; 0 = no preference; 1 = completely right; 0.75 = cutoff
between dextral and adextral). They also received the Lindamood test of
Auditory Conceptualization (LAC) (Lindamood and Lindamood, 1979).
The LAC assesses the metalinguistic ability to conceptualize the
phonemic structure of language by the manipulation of colored blocks
that symbolize phonemes. Performance is scored on a scale of 0–100.
Scaled or normalized scores are not calculated but norms are provided.
Children in the sixth grade (11 years old) are expected to score 100%. In
addition, most of the children received a neurolinguistic battery (Gauger
et al., 1997) and a test of nonverbal intelligence (TONI) (Brown et al.,
1982). An attempt was made to obtain children with a range of
socioeconomic status (SES) (Hollingshead, 1975). Most adults were
recruited to match the parental education of a sample of male patients
with schizophrenia and received the Wechsler Adult Intelligence Scale
(Wechsler, 1987). The remainder of the adult sample were college
students. Some demographic characteristics of the sample are given in
Table 1. The samples are well matched on nonverbal IQ but the children
are more right-handed (P < 0.05) than the adults. Reanalysis of the data
after matching the two groups on handedness did not change any of the
findings reported here.
Figure 2. A common stem duplication caused by a sulcus intermedius (SI) in sagittal and axial planes of section at x = 38 in the left hemisphere. Grey hairlines indicate planes of
intersection. (a) SI does not extend to the base of H. (b) Arrowheads indicate medial and lateral extent of SI. (c) Magnification of (b). (d) Magnification of (a).
MRI Scan
The scans were performed in a Siemens 1 or 1.5T Magnetom (quadrature
head coil) with a three-dimensional MPRage ‘Turbof lash’ technique. Scan
parameters: repetition time (TR),10 ms; echo time (TE), 4 ms; f lip angle,
10o; field of view (FOV), 25 cm; matrix = 130 × 256. The series were
gapless sequences of 1.25 or 1.4 mm slices. The images were transferred
to a computer workstation where they were rotated into the Talairach
plane (horizontal anterior commissure–posterior commissure axis) and
reformatted into 1.00 mm sagittal images using programs written in
PVWave (Visual Numerics, Boulder CO) (Leonard et al., 1993, 1996).
procedure corrects for differences in brain size, without warping or
distorting individual brain anatomy.
Localization in Talairach Space
Parameter files were created for each scan that contained the distance
between the anterior commissure and the edges of the brain in all three
Talairach planes. The Talairach procedure was used to normalize the
identified locations (Talairach and Tournoux, 1988). This procedure
involves dividing the actual location in millimeters by Talairach distances
obtained from the brain measurements in the Talairach atlas. This
Major Landmarks
Figure 1c depicts the sulcal locations that were located in all hemispheres:
(1) HA, the anterior border of the transverse gyrus of Heschl; (2) H1S, the
posterior border of Heschl’s gyrus; and (3) PHRP, the termination of
the posterior horizontal ramus (PHR) that forms the posterior border
of the planum temporale. In cases where there was a terminal posterior
ascending ramus, PHRP was located at the origin of this ramus. When
Landmark Identification
For each hemisphere, 12 images 2 Talairach mm apart between 34 and 56
lateral to the midsagittal plane (x-coordinate) were displayed simultaneously. A cursor was used to mark the depth of five sulcal locations in
each image where they could be identified. The y (horizontal) and z
(vertical) position of the cursor in each image were converted into
Talairach units and stored in an ASCII file for later analysis.
Cerebral Cortex Jul/Aug 1998, V 8 N 5 399
Figure 3. A posterior duplication (PD) in coronal, sagittal and axial planes of section. Grey hairlines indicate planes of intersection. (a) H and H2 cannot be seen simultaneously in
coronal plane. (b) In the sagittal plane there is a clear separation between H1 and H2. Arrowhead indicates position of H2S. (c) Axial plane. Arrowhead indicates medial border of H2S.
there was no ascending ramus, PHRP was marked at the most posterior
extent of PHR on all sections where the fissure extended posteriorly to
H1S.
Landmarks Associated with Heschl Duplications
Two additional sulci were marked when they appeared: (i) SI — the sulcus
that forms a CS duplication by indenting the crown of H, causing it to
develop a heart-shaped appearance (Fig. 2). SI varies considerably in
lateral location and extent. (ii) H2S, the posterior border of a second H
associated with a complete PD. This sulcus was identified when a
separate gyrus emerged lateral to the insula, just caudal to H1S (see Fig.
400 Location of Auditory Cortex Landmarks • Leonard et al.
3). Distinguishing between a CS duplication and a PD in the sagittal plane
is sometimes difficult, especially when an SI is found at medial positions.
A CS duplication can generally be seen in sagittal and coronal planes, but
is sometimes not deep enough to be visible in the axial plane. By contrast,
a complete PD can be identified in sagittal and axial planes but cannot be
visualized in the coronal plane because the second Heschl’s gyrus occurs
completely posterior to the first (compare Figs 2 and 3). The morphology
of the two formations is also typically different. Initially, SI does not
extend to the base of H and the two divisions remain connected with a
heart-shaped or bow-tie appearance. A PD, by contrast, forms a fully
separate structure shaped like a rectangle or a square.
increases. Some landmarks disappear while others remain
constant for at least 2 cm. Figure 5 shows the proportion of
hemispheres in which each landmark was present for each
sagittal position. Since there were few differences between
adults and children, the general features of each sulcus will first
be described in adults.
The Sulcus Anterior to the First Transverse Gyrus of Heschl
(HA)
This sulcus is the first of the five to appear at the postero-lateral
corner of the insula (x = 34). In all 105 hemispheres examined it
continued anterolaterally for at least 1 cm (Fig. 4). It disappeared
more rapidly in the right hemisphere than the left. At x = 56 HA
remained in 40% of the left and 20% of the right hemispheres
(Fig. 5). The HA shifted 20 mm anteriorly and dropped 4 mm
during its course anteriorly on both left and right. From x = 34 to
x = 54, HA was found more anteriorly on the right (paired t-tests,
P < 0.0001; effect sizes ranging from 0.9 to 1.5).
Figure 4. Graphical representation of the course of the major sulci located in this study.
Top: Means and standard deviations for right (negative x) and left HA; the sulcus the
forms the anterior boundary of Heschl’s gyrus. Bottom: means and standard deviations
for H1S, the sulcus forming the posterior boundary of Heschl’s gyrus and PHRP, the
posterior termination of the sylvian fissure. The space between H1S and PHRP is filled
by the planum temporale. The asymmetry of the planum increases with distance from
the midline due to the diverging course of PHRP on left and right.
Two operators marked the landmarks in all 105 brains. The means for
most landmarks differed by <2 mm between the two raters. Initial
intraclass reliability coefficients ranged from 0.93 for HA to 0.5 for medial
positions of SI. In cases of disagreement discussions were held in order to
develop a system of rules and landmarks were then reassessed.
Data Analysis
The files with Talairach locations were analyzed with analysis of variance
and regression methods using PC SAS. Means and standard deviations
were calculated as well as medial–lateral differences in all coordinates.
Distributions of each statistic were plotted by age, sex and handedness
groups as well as for each type of duplication. Analyses of variance with
post hoc t-tests with Bonferroni correction for multiple comparisons were
run to test whether age, sex or handedness contributed significantly to
variance in the position of any of the sulci or to their asymmetr y.
Correlations between landmark position, quantitative handedness
quotient and age were also calculated.
Results
This study had three objectives: to measure (i) variation in the
location and frequency of the anterior and posterior borders of
Heschl’s gyrus and the planum temporale in Talairach space; (ii)
the frequency of the two types of duplications of Heschl’s gyrus;
and (ii) the inf luence of age, sex and handedness on the position
and frequency of these landmarks.
Location in Talairach Space
The major auditory landmarks all move anteriorly (Fig. 4) and
shift slightly inferior (Table 2) as distance from the midline
The Sulcus Posterior to the First Transverse Gyrus of Heschl
(H1S)
A prominent sulcus marking the posterior border of H was found
in all hemispheres by x = 38. There was a slight but significant
lateral asymmetry in the position (left more posterior). The
sulcus coursed anteriorly at about a 45° angle in both
hemispheres and was still present at x = 56 in 92% of the left and
72% of the right hemispheres (Fig. 5). It shifted 8 mm inferiorly
during its anterior course. From x = 34 to x = 54, H1S was found
more anteriorly on the right (paired t-tests, x = 34 to x = 46, P <
0.0001; x = 48, P < 0.001; x = 50, P < 0.01; x = 52 to x = 54, P <
0.05; effect sizes = 2.1 to 0.9 from x = 34 to x = 42: effect sizes
<0.6 lateral to x = 44 ).
The Posterior Termination of the Sylvian Fissure (PHRP)
This landmark was visible in all hemispheres by x = 44. It was the
only landmark that diverged during its lateral course in the two
hemispheres (Fig. 4). It moved posteriorly by 5 mm in the left
and anteriorly by 6 mm in the right. Thus a modest lateral
asymmetry at medial positions increased substantially between x
= 34 and x = 56. This asymmetry was significant at all lateral
positions (paired t-test, P < 0.0001; effect sizes = 0.7–1.7).
Between x = 48 and x = 50 this landmark was further posterior
on the left in 50/53 adults. In contrast, there was no lateral
asymmetr y in the dorsoventral coordinates at any lateral
position.
Landmarks Associated with Heschl’s Duplications: Sulcus
Intermedius
The average AP position of SI moved anteriorly ∼15 mm between
x = 34 and 56 and there was no lateral asymmetry in average
position or anterior course (Fig. 6). The frequency of an SI rose
and then dropped with distance from the midline (Fig. 7). In fact
the frequency quadrupled between x = 34 and x = 42 in each
hemisphere. A SI was found between x = 43 and x = 51 in 40% of
the left and 50% of the right hemispheres. By x = 56, SI was only
visible in 21% of the left and 32% of the right hemispheres.
Complete Posterior Duplications: H2S
A separate gyrus caudal to the first transverse gyrus was found in
seven (13%) of the left and eight (15%) of the right hemispheres.
In five of the left and six of the right hemispheres this gyrus was
still visible at x = 56. In contrast to the sulci described above, H2S
Cerebral Cortex Jul/Aug 1998, V 8 N 5 401
Table 2
y and z coordinates of sulci associated with Heschl’s gyrus and the planum temporale at x = 38 and x = 48 (mean and SD)
x
Children
Adults
Left
Right
z
Left
z
Right
n
y
n
y
n
y
n
y
HA
38
48
52
47
–25.7 ± 3.5
–14.1 ± 5.3
5.9 ± 2.7
4.6 ± 3.0
52
39
–21.7 ± 3.4
–8.8 ± 4.7
6.0 ± 3.0
3.5 ± 3.2
53
48
–26.3 ± 3.8
–12.3 ± 7.1
z
4.4 ± 2.3
1.9 ± 4.1
53
45
–21.0 ± 4.3
–5.8 ± 5.9
z
H1S
38
48
53
53
–32.4 ± 4.6
–24.1 ± 7.3
10.1 ± 3.5
7.8 ± 3.4
53
53
–27.9 ± 4.8
–19.7 ± 7.5
11.4 ± 3.9
7.7 ± 5.0
53
53
–33.6 ± 4.5
–23.0 ± 6.1
9.8 ± 3.0
5.9 ± 3.4
53
53
–29.3 ± 4.5 11.5 ± 3.4
–19.7 ± 6.3 6.9 ± 3.7
PHRP
38
48
49
52
–40.1 ± 4.2
–43.2 ± 6.0
19.7 ± 4.7
22.7 ± 5.7
49
52
–35.5 ± 3.7
–33.5 ± 5.3
21.1 ± 3.8
20.9 ± 5.3
45
53
–41.2 ± 5.3
–44.0 ± 6.5
18.1 ± 4.8
21.1 ± 5.4
47
53
–35.5 ± 4.6 18.5 ± 3.9
–33.1 ± 6.6 19.4 ± 5.3
SI
38
48
16
29
–28.1 ± 3.2
–19.9 ± 4.8
12.0 ± 3.4
9.4 ± 2.8
14
29
–23.9 ± 4.0
–16.7 ± 4.6
11.5 ± 2.8
8.9 ± 3.4
5
22
–31.4 ± 2.4
–18.8 ± 7.3
12.4 ± 4.0
7.5 ± 3.0
12
29
–25.2 ± 3.7 11.9 ± 2.3
–14.2 ± 5.3 6.8 ± 3.3
H2S
38
48
8
10
–39.9 ± 2.6
–34.5 ± 3.6
14.9 ± 4.3
11.8 ± 2.9
6
8
–35.6 ± 3.1
–29.6 ± 4.5
18.4 ± 2.7
15.4 ± 5.2
5
6
–43.0 ± 5.4
–42.6 ± 4.8
17.1 ± 1.6
15.8 ± 2.4
7
8
–34.0 ± 5.2 18.0 ± 4.2
–29.0 ± 6.5 14.1 ± 5.5
3.8 ± 3.4
1.4 ± 3.5
,
frequently remained anchored in a posterior position in the left
hemisphere and did not take an anterolateral course (Fig. 6).
Duplication of Heschl’s Gyrus
The frequency of duplication depended on distance from the
midline (Fig. 7). Duplications caused by a SI indenting the crown
were more frequent than those caused by a complete PD.
Duplications occurred with approximately equal frequency in
each hemisphere in children but were slightly more frequent in
the right hemisphere in adults. As noted in the Introduction,
previous reports of the frequency of duplication of Heschl’s
gyrus have varied widely (Table 3). Since previous studies did
not distinguish between the incidence of CS duplications and
complete PDs, we combined the two types for the purposes of
comparison with previous work. As can be seen from the bottom
half of Table 3, the incidence of duplications increased with
distance from the midsagittal plane. At x = 34 , 73% of the brains
had single gyri bilaterally, but by x = 46, only 17% of the
brains had single gyri bilaterally. [The incidence of duplications
was not tabulated more laterally because of landmark instability (see Fig. 4)]. At each lateral position the distribution of
frequencies agrees with one of the previous studies. The fact
that the number of sulci associated with Heschl’s gyri varies
with lateral position could explain the wide range of Heschl
duplications reported in the literature. Different groups of
investigators may have used different subjective criteria for the
identification of a second Heschl’s gyrus.
At lateral positions, the incidence of CS duplications was
much greater than that of complete PDs. An attempt was made to
determine whether the frequency of different combinations was
different from that expected from random assortment of the
different types. This analysis suggested that a left CS/right PD
might occur less often than expected by chance while a right
PD/left CS might occur more often. The difference between
expected and obser ved frequencies was not significant,
however, because of the large number of cells with low expected
values.
Effect of Demographic Variables
Although only two-thirds of the group fell within 1 cm of the
average position of any sulcus, little of this variance could be
attributed to sex or handedness. None of the sulcal positions
correlated significantly with the handedness quotient. When
handedness was treated as a categorical variable, either as
writing hand or by use of the arbitrary cutoff score of 0.75,
402 Location of Auditory Cortex Landmarks • Leonard et al.
group means were always within a few millimeters of each other
with similar standard deviations and no significant differences.
There were also no significant differences with regard to sex.
The sample size was not large enough to test for interactions
between handedness and sex.
Age
Figures 4–8 and Table 2 demonstrate remarkably few developmental differences in the coordinates of the major sulci. The
most striking difference between the children and adults was in
the anterior trajectory of PDs on the left. In adults, there was no
mean AP shift in H2S while in children, the mean position of the
sulcus moved forward 5 mm. In children, duplications occurred
with approximately equal frequencies in the two hemispheres,
while in adults, CS duplications were more frequent in the right
hemisphere. This lateral asymmetry in duplication in adults was
only significant at x = 44, however. A longitudinal investigation
will be necessary to determine whether the absence of
significant developmental differences is real or due to the
variance inherent in a cross-sectional design.
Discussion
The identification of sulcal landmarks is not straightforward
(Ono et al., 1990; Paus et al., 1996). Although the general
trajectories of major sulci such as the cingulate sulcus, central
sulcus and superior temporal sulcus are easily recognized,
interruptions and branches complicate the identification of
comparable locations in different brains. The instability of sulcal
landmarks poses difficulties for the measurement of cortical
volumes with parcellation methods (Caviness et al., 1996). Novel
methods for visualization of human cortical structure may
require novel approaches to quantification. One novel method
that has been successfully applied to volumetric MRIs is the
brain averaging technique (Evans et al., 1992). In a variant of
that approach Penhune et al. (1996) and Paus et al. (1996) have
described the location of Heschl’s gyrus and the cingulate sulcus
in probability space. As an extension and partial replication of
the Penhune study, we have used a slightly different technique to
describe normal variation in the location and frequency of the
major and minor sulci associated with Heschl’s gyrus.
The study had three objectives: to determine (i) the variability
in standardized three-dimensional space of the major auditory
cortex landmarks; (ii) the incidence of Heschl duplications in the
left and right hemispheres; and (iii) the inf luence of demo-
Figure 5. Landmark probability as a function of distance from the midline. In both
children (top) and adults (bottom) HA and H1S disappear more medially on the right
than the left. All three landmarks are present simultaneously in all brains in a small
region near x = 42.
graphic and cognitive factors such as age, sex and handedness on
landmark frequency, location and asymmetry. The major results
were that (i) there was a 3 cm range in the normalized location
of each sulcal landmark; (ii) in spite of this variability, there were
consistent left–right asymmetries in landmark position; (iii) the
incidence of duplication increased with distance from the
midline; and (iv) the effects of age, sex and handedness were not
powerful enough to contribute significantly to variance in sulcal
frequency, position or asymmetry.
In a pioneering but smaller study of 40 brains, Penhune et al.
(1996) also found a 3 cm variation in the boundaries of Heschl’s
gyrus accompanied by a leftward asymmetry in the volume of
the white matter. The average minimum and maximum A P
coordinates reported in their paper correspond remarkably
well with those reported here, given the differences in sample
characteristics, scan parameters and image-processing techniques between the two studies. This agreement provides
confidence that in future investigations, it will be possible to
characterize individual Heschl’s gyri in terms of their position
with respect to a population rather than their deviance from an
atlas mean.
What Is the Significance of Individual Variation in
Landmark Position?
Although individual variability in sulcal landmarks is sometimes
interpreted as anatomical noise, it is equally possible that
variation in gyral size and position is a consequence of
differences in neural connectivity (Van Essen, 1997). In support
of the functional significance of gyral size, previous studies have
Figure 6. Graphical representation of the course of the sulci associated with Heschl
duplications. Top: Means and standard deviations for right (negative x) and left sulcus
intermedius (SI), a sulcus that indents Heschl’s gyrus to form a common stem
duplication. Bottom: means and standard deviations for H2S, the sulcus that forms the
posterior boundary of a second complete Heschl’s gyrus. The left H2S does not angle
forward as sharply as that on the right, particularly in adults.
found that women and left-handers have less planar asymmetry
(Steinmetz et al., 1991; Witelson and Kigar, 1992; Foundas, et
al., 1995). In the present study, neither sex nor handedness
accounted for significant variance in either the anterior or
posterior boundary of the planum. Methodological or sampling
differences may account for our failure to find sex and
handedness effects.
It is remarkable how much effort has been devoted to the
search for anatomical differences due to sex and handedness in
contrast to those underlying cognitive and sensory processing
differences. Behaviors based on auditory input — verbal and
nonverbal communication, music and sound localization — have
been of critical importance in the evolution of our species
(Hauser, 1996). In order for a behavior to evolve through natural
selection, it must vary. Auditory and language function do
indeed vary markedly among individuals (Bates et al., 1995).
Dysregulated variation in the neurobiological substrates for
these functions has been hypothesized as the cause of behavioral
disorders as different as dyslexia (Liberman et al., 1974;
Liberman and Shankweiler, 1985), specific language impairment
(Elliot et al., 1989; Merzenich et al., 1996; Tallal et al., 1996) and
schizophrenia (Crow, 1990). When functional variation has
arisen through natural selection, anatomical substrates for this
variation can generally be identified in animals. Classic examples
of form ref lecting function are found in the cortical ‘barrel
fields’ (Woolsey and Van der Loos, 1970) and the ocular
dominance columns in cat and monkeys (Hubel et al., 1977).
The size and arrangement of these arrays are dictated genetically
but can be altered by manipulations of sensory experience. An
Cerebral Cortex Jul/Aug 1998, V 8 N 5 403
Table 3
Comparison of frequency of different right–left Heschl combinations (adapted from Penhune et al.,
1996)
Type of
material
Study
n
Frequency (%)
No. of
L
Heschl’s gyri R
Postmortem
Postmortem
Postmortem
Postmortem
MRI
MRI
Von Economo and Horn (1930)
Campaign and Minkler (1976)
Musiek and Reeves (1990)
Rademacher et al. (1993)
Penhune et al. 1 (1996)b
Penhune et al. 2 (1996)
present study
a
11
30
29
10
20
20
105
x
34
36
38
40
42
44
46
1
1
1
2
2
1
2
2
18
10
17
40
70
60
55
48
20
30a
20
10
18 9
3 36
24 34
30 0
5 5
25 5
70
55
42
32
23
17
16
13
21
26
29
30
29
28
15
19
22
24
19
14
16
2
5
10
15
28
40
40
Figure includes one case with a left single, right triple.
b
Penhune treated their two samples of 20 subjects as two separate studies. In the present study,
the incidence of duplications changes as a function of laterality (x).
Figure 7. Probability of Heschl duplication as a function of distance from the midline.
The probability of common stem duplications peaks between x = 44 and x = 48 on
both left and right sides. The probability of posterior duplications rises between x = 34
and x = 40 and then remains constant at slightly under 0.2 in both hemispheres.
example of a genetically programmed anatomical substrate
whose development is epigenetically regulated by auditory input
is the vocal control network of the songbird (Nottebohm et al.,
1976; Bottjer and Johnson, 1992). A network of nuclei ‘visible to
the naked eye’ is found in the sex and species that learn their
song and is reduced in size or absent in females and species with
stereotyped song. Manipulation of experience and hormonal
environment can shrink or expand nuclei in the network with
dramatic functional effects.
These animal model systems provide the scientific foundation
for speculating that variation in anatomical size and shape of
human brain structures can have functional significance. The
parallel to the animal work is clear. Somatosensory information
provided by the whiskers dominates the world of the mouse.
Humans, on the other hand, depend on oral language to
interpret experience. The sensor y system necessar y for the
normal development of oral language is audition. Instead of
barrel fields, humans have a prominent bulge in the lower bank
of the sylvian fissure: Heschl’s gyrus. This suggests the testable
hypothesis that variation in the size, shape and asymmetry of
Heschl’s gyrus and the planum ref lects variation in the size of
auditory receptive fields, their underlying connectivity and
accompanying auditory detection performance.
In a recent article, Gannon et al. (1998) suggest that human
planar asymmetry may be an ‘epiphenomenon’ (p. 221) because
chimpanzees also have planar asymmetry but do not depend on
oral language for communication. They suggest that planar
asymmetry could underlie diverse abilities in the two species.
Tallal et al. (1996) have proposed that the left hemisphere in a
variety of mammals is specialized for discrimination of very
404 Location of Auditory Cortex Landmarks • Leonard et al.
short sounds. In some animals this ability might be used for prey
location, in others for mate recognition, while in humans it
might facilitate the discrimination of consonants. In any species,
individual differences in the ability to perform this type of
discrimination might correlate with epigenetically regulated
differences in size and shape.
What Is the Base Rate of Heschl Duplications?
The incidence of Heschl’s duplications has been reported to be
elevated in individuals with dyslexia (Leonard et al., 1993) and
resistance to thyroid hormone (Leonard et al., 1995). In both of
these studies the identification of Heschl duplications was
restricted to sections just medial to the insula. At medial
positions, the incidence of such duplications is low in the
normal adults studied here. The fact that the incidence of medial
duplications is higher in the children than in the adults is
difficult to interpret in this cross-sectional study. It is possible
that there are developmental changes in Heschl morphology, but
it seems more reasonable to suppose that there are sampling
differences in our child and adult population. We are presently
engaged in a longitudinal study to distinguish between these two
interpretations.
This study provides a possible resolution to one question that
has occasioned considerable debate — the prevalence of Heschl
duplications in normal adults. Because postmortem studies have
reported a high incidence of duplications, the significance of an
elevation in clinical populations has been difficult to evaluate. A
recent quantitative MRI study of normal subjects (Penhune et al.,
1996), however, found the incidence to be only ∼20%. Our study
suggests that range of frequencies in previous reports may
ref lect a range of criteria for duplication. It turns out that the
landmarks that define the duplications are not constant but
emerge and disappear at varying lateral positions. The frequency
of duplication rises and falls with distance from the midline.
It is time to recognize the arbitrary nature of landmark
identification and the relatively abstract, derived and evanescent
nature of concepts such as Heschl’s gyrus and Heschl’s duplications. The subtitle of this paper is a paraphrase of Bogen and
Bogen (1976). Their paper was a scholarly rumination on the
lack of overlap between depictions of Wernicke’s area found in
various sources. Since Wernicke’s area is a concept, it should not
be so surprising that it would have many cortical instantiations.
Anatomical landmarks, however, are supposed to be objects, not
concepts. They should exist in a particular location. On medial
sections, this expectation was fulfilled for Heschl’s gyrus. It was
easily visualized in all 105 brains. Laterally, however, the
boundary landmarks faded out gradually. Caviness and his
colleagues have solved the problem of landmark instability with
canonical planes and landmark boundaries in the hope that other
laboratories will adopt a common set of criteria for cortical
parcellation (Rademacher et al., 1993; Caviness et al., 1996). But
the location of these canonical planes and sulci has been
investigated in a relatively small number of brains. The validity of
these parcellation techniques is compromised when brains are
encountered which vary so far from the canonical type that the
rules become difficult, if not impossible, to apply (Steinmetz et
al., 1990a; Witelson and Kigar, 1992; Ide et al., 1996). Our
studies have adopted an alternate strategy and established
boundary rules that exaggerate the difference between normal
and patient populations (e.g. by restricting the definition of
Heschl’s duplications to medial slices where the frequency is low
in normal adults). Ideally, of course, structure boundaries should
be defined by function (King et al., 1997). Appropriately
designed functional imaging experiments that map signal
elevations onto sulcal banks in single subjects (Crosson et al.,
1998; Gokcay et al., 1998) may reveal less arbitrary signposts to
the construction of a functionally valid anatomy.
Notes
We thank Samuel Browd, Laurie Mercado-Gauger, Melissa Mahoney, Erin
Gautier, Tim Lucas, Marina Santarpia, Mark Eckert, Lisa Rowe, Paula
Zuffante, Dawn Heron, Wendy Swearingen, Mar y Ellen Bentham,
Christine Meyer and Bill Bell for their indispensable contributions. We
also thank Patrick Barta for providing a prepublication copy of Measure.
This research was supported by Social and Behavioral Sciences Research
Grant no. 12 FY93–0551 from the March of Dimes Birth Defects
Foundation, the Research Service of the Veterans’ Administration, grant
RO1 DC 02292 from the National Institute of Deafness and Communication Disorders and a National Science Foundation Collaboration
Research Initiative Grant.
Address correspondence to Christiana M. Leonard, PO Box 100244,
Department of Neuroscience, University of Florida Brain Institute,
Gainesville, FL 32610, USA. Email: [email protected] l.edu.
References
Allman J (1987) Maps in context: some analogies between visual cortical
and genetic maps. In: Matters of intelligence (Vaina L, ed), pp.
369–393. Dordrecht: Reidel.
Allman J, Kaas J (1971) A representation of the visual field in the caudal
third of the middle temporal gyrus of the owl monkey (Aotus
trivirgatus). Brain Res 31:85–101.
Bates E, Dale P, Thal D (1995) Individual differences and their
implications for theories of language development. In: Handbook of
child language (Fletcher P, MacWhinney, B eds), pp. 96–163. Oxford:
Blackwell.
Bogen J, Bogen G (1976) Wernicke’s region: where is it? Ann NY Acad Sci
280:834–843.
Bookstein F (1996) Biometrics, biomathematics and the morphometric
synthesis. Bull Math Biol 58:313–365.
Bottjer S, Johnson F (1992) Matters of life and death in songbird forebrain.
J Neurobiol 23:1172–1191.
Briggs G, Nebes R, Kinsbourne M (1976) Intellectual differences in
relation to personal and family handedness. Quart J Exp Psychol
28:591–601.
Brown L, Sherbenau R, Johnsen S (1982) Test of nonverbal intelligence.
Austin, TX: Pro-Ed.
Campain R, Minckler J (1976) A note on the gross confiurations of the
human auditory cortex. Brain Lang 3:318–323.
Caviness V Jr, Meyer J, Makris N, Kennedy D (1996) MRI-based
topographic parcellation of human neocortex: an anatomically
specified method with estimate of reliability. J Cognit Neurosci
8:566–587.
Crosson B, Sadek JR, Bobholz JA, Gokcay D, Mohr C, Freeman AJ,
Maron L, Auerback EJ, Leonard CM, Berg S, Browd S, Briggs RW
(1998) Precise location of medial frontal activity during word
generation by fMRI. Neuroimage (in press).
Crow T (1990) Temporal lobe asymmetries as the key to the etiology of
schizophrenia. Schizophr Bull 16:443–443.
Elliot L, Hammer M, Scholl M (1989) Fine-grained auditory discrimination
in normal children and children with language learning problems. J
Speech Hear Res 32:112–119.
Evans A, Marrett S, Neelin P, Collins L, Worsley K, Dai W, Milot S, Meyer
E, Bub D (1992) Anatomical mapping of functional activation in
stereotactic coordinate space. Neuroimage 1:43–53.
Felleman D, Van Essen D (1991) Distributed hierarchical processing in
the primate cerebral cortex. Cereb Cortex 1:1–47.
Foundas A, Leonard C, Heilman K (1995) Morphological cerebral
asymmetries and handedness: the pars triangularis and planum
temporale. Arch Neurol 52:501–508.
Galaburda A (1989) Ordinar y and extraordinar y brain development:
anatomical variation in developmental dyslexia. Ann Dyslexia 39:
67–79.
Gannon P, Holloway R, Broadfield D, Braun A (1988) Asymmetr y of
chimpanzee planum temporale: humanlike pattern of Wernicke’s
brain language area homolog. Science 279:220–222.
Gauger L, Lombardino L, Leonard C (1997) Brain morphology in children
with specific language impairment. J Speech Lang Res 40:1272–1284.
Gokcay D, Mohr C, Crosson B, Leonard CM, Bobholz JA (1998) Software
for ROI analysis of fMRI images based on anatomic landmarks.
Neuroimage (in press).
Geschwind N, Levitsky W (1968) Human brain: left–right asymmetries in
temporal speech region. Science 161:186–187.
Hauser M (1996) Evolution of human communication. Cambridge, MA,
Harvard University Press.
Hollingshead A (1975) Four factor index of social skills. New Haven, CT:
Yale University Press.
Hubel D, Wiesel T, LeVay S (1977) Plasticity of ocular dominance columns
in monkey striate cortex, Phil Trans R S 278:377–409.
Ide A, Rodriguez E, Zaidel E, Aboitiz F (1996) Bifurcation patterns in the
human sylvian fissure: hemispheric and sex differences. Cereb Cortex
6:717–725.
King W, Mohr C, Freeman A, Leonard C (1997) Effect of stimulus intensity
on auditory cortex using fMRI. Soc Neurosci Abstr 23.
Kulynych J, Vladar K, Jones D, Weinberger D (1994) Gender differences
in the normal lateralization of the supratemporal cortex — MRI
surface-rendering morphometry of Heschl’s gyrus and the planum
temporale. Cereb Cortex 4:107–118.
Leonard C, Martinez P, Weintraub B, Hauser P (1995) Magnetic resonance
imaging of cerebral anomalies in subjects with resistance to thyroid
hormone. Am J Med Genet 60:238–243.
Leonard C, Lombardino L, Mercado L, Browd S, Breier J, Agee O (1996)
Cerebral asymmetr y and cognitive development in children: A
magnetic resonance imaging study. Psychol Sci 7:79–85.
Leonard C, Voeller K, Lombardino L, Morris M, Alexander A, Andersen H,
Garofalakis M, Hynd G, Honeyman J, Mao J, Agee O, Staab E (1993)
Anomalous cerebral structure in dyslexia revealed with magnetic
resonance imaging. Arch Neurol 50:461–469.
Liberman A, Shankweiler D (1985) Phonology and the problems of
learning to read and write. Remedial Special Educ 6:8–17.
Liberman I, Shankweiler D, Fischer F, Carter B (1974) Explicit syllable and
phoneme segmentation in the young child. J Exp Child Psychol
18:201–212.
Liegeois-Chauvel C, Musolino A, Chavel P (1991) Localization of the
primary auditory area in man. Brain 114:139–153.
Lindamood C, Lindamood P (1979) Lindamood auditory conceptualization test. Boston, MA: Teaching Resources Corporation.
Mercado L, Browd S, Leonard C, Lombardino L, Voeller K, Ross J,
Noffzinger K, Agee O (1994) Sylvian fissure anomalies in children
with learning disabilities. Soc Neurosci Abstr 22:1481.
Merzenich M, Brugge J (1973) Representation of the cochlear partition of
the superior temporal plane of the macaque monkey. Brain Res
50:275–276.
Merzenich M, Jenkins W, Johnston P, Schreiner C, Miller S, Tallal P (1996)
Cerebral Cortex Jul/Aug 1998, V 8 N 5 405
Temporal processing deficits of language-learning impaired children
ameliorated by training. Science 271: 77–81.
Musiek F, Reeves A (1990) Asymmetries of the auditor y areas of the
cerebrum. J Am Acad Audiol 1:240–245.
Nottebohm F, Stokes T, Leonard C (1976) Central control of song in the
canary. J Comp Neurol 165:457–486.
Ono M, Jubik S, Abernathy C (1990) Atlas of the cerebral sulci. New York:
Thieme.
Paus T, Tomaiuolo F, Otaky N, MacDonald D, Petrides M, Atlas J, Morris R,
Evans A (1996) Human cingulate and paracingulate sulci; pattern,
variability, asymmetr y, and probabilistic map. Cereb Cortex 6:
207–214.
Penhune V, Zatorre R, MacDonald J, Evans A (1996) Interhemispheric
anatomical differences in human primary auditory cortex; probabilistic mapping and volume measurement from magnetic resonance
scans. Cereb Cortex 6:661–672.
Rademacher J, Caviness V, Steinmetz H, Galaburda A (1993) Topographical variation of the human primary cortices; implications for
neuroimaging, brain mapping and neurobiology. Cereb Cortex
3:313–329.
Rauschecker J, Tian B, Hauser M (1995) Processing of complex sounds in
the macaque nonprimary auditory cortex. Science 268:111–114.
Schlaug G, Jancke L, Huang Y, Steinmetz H (1995) In vivo evidence of
structural brain asymmetry in musicians. Science 267: 699–701.
Sereno M, Dale A, Tootell R, Reppas J, Kwong K, Belliveau J, Brady T,
Rosen B (1995) Human visual areas identified by visual field sign on
the cortical surface using phase-encoded retinotopic stimulation.
Science 268:889–893.
Steinmetz H, Rademacher J, Huang Y (1989) Cerebral asymmetry: MR
planimetry of the human planum temporale. J Comput Assist Tomog
13:996–1005.
Steinmetz H, Ebeling U, Huang Y, Kahn T (1990a) Sulcus topography of
the parietal opercular region: an anatomic and MR study. Brain Lang
38:515–533.
406 Location of Auditory Cortex Landmarks • Leonard et al.
Steinmetz H, Rademacher J, Jancke L, Huang Y, Thron A, Zilles K (1990b)
Total surface of temporoparietal intrasylvian cortex; diverging left–
right asymmetries. Brain Lang 39:357–372.
Steinmetz H, Volkmann J, Jancke L, Freund H-J (1991) Anatomical left–
right asymmetry of language related temporal cortex is different in
left- and right-handers. Ann Neurol 29:315–319.
Talairach J, Tournoux P (1988) Coplanar stereotaxic atlas of the human
brain: three-dimensional proportional system: an approach to
cerebral imaging. New York: Thieme.
Tallal P, Miller S, Fitch RH (1993) Temporal processing in the nervous
system; implications for the development of phonological systems.
Ann NY Acad Sci 682:27–47.
Tallal P, Miller S, Bedi G, Byma G, Wang X, Nagarajan S, Schreiner C,
Jenkins W, Merzenich M (1996) Language comprehension in
language-learning impaired children improved with acoustically
modified speech. Science 271:81–84.
Van Essen D (1997) A tension-based theor y of morphogenesis and
compact wiring in the central nervous system. Nature 385:313–318.
Von Economo C, Horn L (1930) Gyral relief, size and cortical architectonics of the suprotemporal surface: their individual and lateral
differences. Z Ges Neurol Psychiat 130:687–757.
Wechsler D (1987) Wechsler Memor y Scale — revised manual. San
Antonio, TX: Psychological Corporation.
Witelson S (1982) Bumps on the brain: right–left asymmetries in temporal
speech region. In: Language functions and brain organization
(Segalowitz S, ed), pp. 117–144. Orlando, FL: Academic Press.
Witelson S, Kigar D (1992) Sylvian fissure morphology and asymmetry in
men and women: bilateral differences in relation to handedness in
men. J Comp Neurol 323:326–340.
Woolsey TA, Van der Loos H (1970) The structural organization of layer IV
in the somatosensory region (SI) of the mouse cerebral cortex: the
description of a cortical field composed of discrete cytoarchitectonic
units. Brain Res 17:205–242.