Neuroscience Letters, 146 (1992) 179-182
179
© 1992 Elsevier Scientific Publishers Ireland Ltd. All rights reserved 0304-3940/92/$ 05.00
NSL 09064
Regional response differences within the human auditory cortex when listening to words Cathy Price a, Richard Wise a'b, Stuart R a m s a y a, Karl Friston a, David H o w a r d c, K a r a l y n Patterson d and Richard Frackowiak a aMRC Cyclotron Unit, Hammersmith Hospital, London (UK), hNeuroscienees Centre, Charing Cross Hospital, London (UK), CDepartmentof Psychology, Birkbeck College, London (UK) and dMRC Applied Psychology Unit, Cambridge (UK) (Received 15 June 1992; Revised version received 5 August 1992; Accepted 7 August 1992)
Key words: PET; Language; Presentation rate; Word; Regional cerebral blood flow; Wernicke's area The relationship between activity within the human auditory cortices and the presentation rate of heard words was investigated by measuring changes in regional cerebral blood flow with positron emission tomography. We demonstrate that in the primary auditory cortices and middle regions of the superior temporal gyri there is a linear relationship between the rate of presentation of heard words and blood flow response. In contrast, the blood flow response in an area of the left posterior superior temporal gyrus (Wernicke's area) is primarily dependent on the occurrence of words irrespective of their rate of presentation. The primary auditory cortices are associated with the early processing of complex acoustic signals whereas Wernicke's area is associated with the comprehension of heard words. This study demonstrates for the first time that time dependent sensory signals (heard words) detected in the primary auditory cortices are transformed into a time invariant output which is channelled to a functionally specialised region - Wernicke's area. Wernicke's area is therefore distinguished from other areas of the auditory cortex by direct observation of signal transformation rather than by association with a specific behavioural task.
Functional anatomy can be studied by measuring changes in regional cerebral blood flow (rCBF) in response to performance on behavioral tasks [10]. The interpretation of these activation studies depends on the relationship between the rate of stimulus presentation, the regional excitatory or inhibitory synaptic activity that underlies the processing of the stimuli and blood flow response. Only one previous study has formally examined the relationship between stimulus rate and rCBF [5]. This looked at the response in primary visual cortex to simple, repetitive photic stimuli and found that rCBF increases were linearly correlated with stimulus rates between 0 and 7.8 Hz. Above 7.8 Hz, rCBF increases plateaued or even fell as stimulus rate increased suggesting that at high stimulus rates neuronal response following each stimulus repetition was no longer uniform. We and many others are interested in language activation studies and frequently use complex auditory stimuli. Knowledge about the rCBF response to heard words in different regions of the auditory cortex is fundamental to such investigations. Several studies have shown that Correspondence: C. Price, MRC Cyclotron Unit, Hammersmith Hospital, Du Cane Road, London W12 OHS, UK.
hearing words or word-like sounds activates posterior temporal regions [14]. The present study investigated the relationship between the rate of presentation of heard words and rCBF increases in the following way. Six right-handed, English speaking, normal, male volunteers, aged 24-49 years, were studied. Each subject gave informed consent to have 6 consecutive measurements of rCBF, using a C1502 inhalation technique [7, 8] and a Siemens 931-08/12 positron emission tomographic scanner [12]. During each 3.5 min dynamic scan, the subject inhaled CI502 at a concentration of 6 MBq/ml and a flow rate of 500 ml/min through a standard oxygen face mask for a period of 2 rain. Intervals between scans lasted 12-15 rain. Correction for attenuation was made by performing a transmission scan with an exposed 68Ge/ 6SGa external ring source at the beginning of each patient study. Images were reconstructed by filtered back projection (Hanning filter, cut off) giving a transaxial resolution of 8.5 mm full width at half maximum. The reconstructed images contained 128 x 128 pixels, each having a size of 2.05 × 2.05 mm. The behavioral state for each measurement was rest or listening to nouns presented at rates of 10, 30, 50, 70 or 90 words per minute (wpm). The subjects were instructed
180 Left superior temporal gyrus (middle part)
74
}t
}
'°'1 ~s
Left Primary
auditory
} a
cortex
b
c
d
e
f
Right Primary auditory
72S
a8 s
temporal gyrus {middle part)
"'
t
690 67O
{ }
'°~
Right superior
cortex
t
}
66~
630 a
Left posterior
superior
temporal
5c
gyrus
b
c
d
Right posterior
e
f
superior
temporal
gyrus
4S s
tt
4as 42s 41 5 a
b
c
~
e
l
}
4o s a
b
c
d
e
f
Fig. 1. Each graph represents adjusted rCBF plotted against task (a = rest, b = 10 wpm, c = 30 wpm, d = 50 wpm, e = 70 wpm, f= 90 wpm) for all subjects in 6 different regions within the auditory cortex. The coordinates of these regions were selected from the atlas of Talairach and Tournoux [13]. For the middle parts of the superior temporal gyri x = _+50,y -- -22, z - 4 mm; R 2 = 0.92 and 0.96, respectively. For the primary auditory cortices x - + 5 0 , y - - 2 2 , z = I2 mm: R 2 - 0.99 and 0.66, respectively. For the posterior parts of the superior temporalgyrix=_+58, y = 3 4 , z - 12 mm; R2 = 0.12 and 0.88, respectively. The linear relationship between increases in rCBF and presentation rate is illustrated in all regions of the auditory cortex with the exception of the left posterior superior temporal gyrus.
to listen to the tapes o f w o r d s with their eyes closed, o r to relax a n d ' e m p t y their m i n d s ' d u r i n g the rest state. Within the different p r e s e n t a t i o n rate conditions, the w o r d sets were m a t c h e d for w o r d frequency, n u m b e r o f syllables a n d r a t e d imageability. The o r d e r o f a c t i v a t i o n c o n d i t i o n s was d e t e r m i n e d by a L a t i n S q u a r e design.
D a t a were a n a l y z e d using statistical p a r a m e t r i c m a p ping ( S P M Software, M R C C y c l o t r o n Unit, U K ) [3]. C a l c u l a t i o n s a n d image m a t r i x m a n i p u l a t i o n s were perf o r m e d in P R O M A T L A B ( M a t h w o r k s Inc., S h e r b o r n , M A , U S A ) . F o r each subject, r C B F t h r o u g h o u t the whole b r a i n region (15 t r a n s a x i a l planes) was c a l c u l a t e d a c c o r d i n g to a previously described p r o t o c o l [8]. In o r d e r to a c c o u n t for i n d i v i d u a l differences in b r a i n size a n d gyral a n a t o m y , the d a t a were r e o r i e n t a t e d parallel to the i n t e r c o m m i s s u r a l line, then n o r m a l i z e d into the stereotactic space o f the a n a t o m i c a l atlas o f T a l a i r a c h and T o u r n o u x [4, 13]. The v o l u m e elements (voxets) o f these n o r m a l i z e d images (2 x 2 x 4 m m in the x, y a n d z d i m e n sions) where then s m o o t h e d with a G a u s s i a n filter 20 m m wide so t h a t voxel values c o r r e s p o n d e d to a weighted m e a n r C B F in a spherical d o m a i n 20 m m in diameter. These s t a n d a r d i s a t i o n processes also p e r m i t the l o c a t i o n o f each voxel to be referenced to the a n a t o m i c a l atlas o f T a l a i r a c h a n d T o u r n o u x [13]. As the study was designed to examine regional changes across activation c o n d i t i o n s , the d a t a were norm a l i z e d for global flow differences by analysis o f covariance ( A N C O V A ) , with m e a s u r e d g l o b a l flow as the covariate, a n d then a v e r a g e d for each c o n d i t i o n across the 6 subjects [2]. This analysis g e n e r a t e d a d j u s t e d voxel m e a n s for each o f the 6 c o n d i t i o n s a n d the associated adjusted e r r o r variance required for their c o m p a r i s o n . The differences between c o n d i t i o n s were assessed by weighting the c o n d i t i o n m e a n s with the a p p r o p r i a t e contrast. This analysis was d o n e for each voxel a n d the resulting set o f t values c o n s t i t u t e d the t-statistical p a r ametric m a p (SPM{ t}). C o m p a r i s o n o f each a u d i t o r y a c t i v a t i o n c o n d i t i o n c o m p a r e d to rest p r o d u c e d highly significant bilateral increases in r C B F involving the m i d d l e and p o s t e r i o r regions o f the s u p e r i o r t e m p o r a l gyri. F o r the slower rates o f p r e s e n t a t i o n , activity in the left p o s t e r i o r s u p e r i o r
Fig. 2. The 3 brain projections viewing the brain from above (transverse), the right (saginal) and the back (coronal) illustrate that the only regions of the brain where rCBF rises linearly with the presentation rate of heard words were the superior and middle temporal gyri (P < 0.05, corrected for non-independent comparisons). This was assessed with statistical parametric mapping by weighting the six condition means with the rate of word presentation.
181
Fig. 3. The 3 brain projections viewing the brain from above (transverse), the right (sagittal) and the back (coronal) illustrate that the only region in the brain where rCBF rose significantly irrespective of presentation rate was the left posterior superior temporal region (P < 0.05, corrected for non-independent comparisons). This was assessed with statistical parametric mapping by identifying the areas of the brain where the rCBF difference between rest and the slowest rate of presentation was significantlygreater than the average rCBF difference between each word presentation rate. The equation applied was [(b-a) - ({f-b}/4)] where a = rest, b = the slowest rate of word presentation and f = the fastest rate of word presentation.
temporal gyrus was m o r e extensive than activity in the h o m o l o g o u s area on the right but this a s y m m e t r y was less a p p a r e n t at faster rates o f presentation. This observation can be accounted for by plotting the m e a n r C B F values for each condition at different regions o f the temporal gyri. Fig. 1 illustrates that in b o t h p r i m a r y auditory cortices and the middle regions o f the superior temporal gyri, r C B F is linearly correlated with presentation rate. In m o r e posterior regions o f the superior temporal gyri, the linear relationship is present on the right but not on the left. In the left posterior superior temporal region, there is an almost u n i f o r m increase in r C B F irrespective o f presentation rate. Fig. 2 displays the regions o f the brain (superior and middle temporal gyri) where r C B F rises linearly with presentation rate. Fig. 3 illustrates that the only region in the brain where r C B F rose significantly irrespective o f presentation rate is in the left posterior superior temporal gyrus (Wernicke's area). Bilateral temporal lobe lesions can result in pure w o r d s o u n d deafness [1] (an inability to decipher heard words while all other language abilities are intact) whereas lesions that include Wernicke's area result in impaired spoken language in addition to a severe deficit in spoken c o m p r e h e n s i o n [11]. The lesion data therefore supports the n o t i o n that early acoustic analysis o f words occurs in b o t h temporal lobes but Wernicke's area is responsible for the further processing o f encoded verbal material. Activation in Wernicke's area has also been d e m o n strated in P E T studies in association with tasks requiring auditory w o r d c o m p r e h e n s i o n [6, 14]. In the present study Wernicke's area was activated even when subjects are not explicitly asked to r e m e m b e r or understand words, a finding consistent with a b u n d a n t psychological evidence that a u d i t o r y w o r d c o m p r e h e n s i o n proceeds automatically [9]. We conclude that input to areas associated with early
acoustic analysis is dependent on the n u m b e r o f auditory stimuli heard per unit time but the o u t p u t f r o m these areas to Wernicke's area results in a response which is independent o f the n u m b e r o f auditory stimuli heard per unit time. Hence, functional specialisation in Wernicke's area is shown on the basis o f signal t r a n s f o r m a t i o n rather than by association with a specific behavioural task. This w o r k was funded by a G r a n t (91-23) f r o m the M c D o n n e l l Pew Program. 1 Auerbach, S.H., Allard, T., Naeser, M., Alexander, M.P. and Albert, M.L., Pure word deafness: analysis of a case with bilateral lesions and a defect at the prephonemic level, Brain, 105 (1982) 271-300. 2 Friston, K.J., Frith, C.D., Liddle, P.F., Dolan, R.J., Lammertsma, A.A. and Frackowiak, R.S.J., The relationship between local and global changes in PET scans, J. Cereb. Blood Flow Metab., 10 (1991) 458466. 3 Friston, K.J., Frith, C.D., Liddle, EF. and Frackowiak, R.S.J., Comparing functional PET images: the assessment of significant change, J. Cereb. Blood Flow Metab., 11 (1991) 690-699. 4 Friston, K.J., Frith, C.D., Liddle, EF. and Frackowiak, R.S.J., Plastic transformation of PET images, J. Comput. Assist. Tomogr., 15 (1991) 634-639. 5 Fox, P.T. and Raichle, M.E., Stimulus rate determines regional brain blood flow in striate cortex, Ann. Neurol., 17 (1985) 303-305. 6 Howard, D., Patterson, K., Wise, R., Brown, W., Friston, K., Weiller, D. and Frackowiak, R., The cortical localisations of the lexicons: PET evidence, Brain, in press. 7 Lammertsma, A.A., Cunningham, M.P., Deiber, J.D., Heather, EM., Bloomfield, RM., Nutt, J., Frackowiak, R.S.J. and Jones, T., Combination of dynamic and integral methods for generating reproducible functional CBF images, J. Cereb. Blood Flow Metab., l0 (1990) 675-686. 8 Lammertsma, A.A., Frackowiak, R.S.J., Hoffman, J.M., Huang, S., Weinberg, I.N., Dahlbom, M., MacDonald, N.S., Hoffman, E.J., Mazziotta, J.C., Heather, J.D., Forse, G.R., Phelps, M.E. and Jones, T., The C1502 build up technique to measure regional cere-
182 bral blood flow and metabolism, J. Cereb. Blood Flow Metab., 9 (1989) 461~170. 9 Marslen-Wilson, W.D. and Tyler, L.K., Against Modularity. In J. Garfield (Ed.), Modularity and Knowledge Representation and Natural Language Understanding, Lawrence Erlbaum, Hillsdale, NJ, 1987. t0 Raichle, M.E., Circulatory and metabolic correlates of brain function in normal humans. In F. Plum (Ed.), Handbook of Physiology, Vol. 5: The Nervous System, Am. Physiol. Soc., Oxford University Press, New York, 1987, 643-674. 11 Seldon, H.L., The anatomy of speech perception: human auditory cortex. In A. Peters and E.G. Jones (Eds.), Cerebral Cortex, Vol. 4: Association and Auditory Cortices, Plenum, New York, 1985.
12 Spinks, T.J., Jones, T., Gilardi, M.C. and Heather, J.D., Physical performance of the latest generation of commercial PET scanner, IEEE (Inst. Electr. Electron. Eng.) Trans. Nucl. Sci., 35 (1988) 721725. 13 Talairach, J. and Tournoux, P., Co-planar Stereotactic Atlas of the Human Brain, Thieme, Stuttgart, 1988. 14 Wise, R.J.S., Chollet, F., Hadar, U., Friston, K., Hoffner, E. and Frackowiak, R.S.J., Distribution of cortical neural networks involved in word comprehension and word retrieval, Brain, 114 (1991) 1803-1817.
179
© 1992 Elsevier Scientific Publishers Ireland Ltd. All rights reserved 0304-3940/92/$ 05.00
NSL 09064
Regional response differences within the human auditory cortex when listening to words Cathy Price a, Richard Wise a'b, Stuart R a m s a y a, Karl Friston a, David H o w a r d c, K a r a l y n Patterson d and Richard Frackowiak a aMRC Cyclotron Unit, Hammersmith Hospital, London (UK), hNeuroscienees Centre, Charing Cross Hospital, London (UK), CDepartmentof Psychology, Birkbeck College, London (UK) and dMRC Applied Psychology Unit, Cambridge (UK) (Received 15 June 1992; Revised version received 5 August 1992; Accepted 7 August 1992)
Key words: PET; Language; Presentation rate; Word; Regional cerebral blood flow; Wernicke's area The relationship between activity within the human auditory cortices and the presentation rate of heard words was investigated by measuring changes in regional cerebral blood flow with positron emission tomography. We demonstrate that in the primary auditory cortices and middle regions of the superior temporal gyri there is a linear relationship between the rate of presentation of heard words and blood flow response. In contrast, the blood flow response in an area of the left posterior superior temporal gyrus (Wernicke's area) is primarily dependent on the occurrence of words irrespective of their rate of presentation. The primary auditory cortices are associated with the early processing of complex acoustic signals whereas Wernicke's area is associated with the comprehension of heard words. This study demonstrates for the first time that time dependent sensory signals (heard words) detected in the primary auditory cortices are transformed into a time invariant output which is channelled to a functionally specialised region - Wernicke's area. Wernicke's area is therefore distinguished from other areas of the auditory cortex by direct observation of signal transformation rather than by association with a specific behavioural task.
Functional anatomy can be studied by measuring changes in regional cerebral blood flow (rCBF) in response to performance on behavioral tasks [10]. The interpretation of these activation studies depends on the relationship between the rate of stimulus presentation, the regional excitatory or inhibitory synaptic activity that underlies the processing of the stimuli and blood flow response. Only one previous study has formally examined the relationship between stimulus rate and rCBF [5]. This looked at the response in primary visual cortex to simple, repetitive photic stimuli and found that rCBF increases were linearly correlated with stimulus rates between 0 and 7.8 Hz. Above 7.8 Hz, rCBF increases plateaued or even fell as stimulus rate increased suggesting that at high stimulus rates neuronal response following each stimulus repetition was no longer uniform. We and many others are interested in language activation studies and frequently use complex auditory stimuli. Knowledge about the rCBF response to heard words in different regions of the auditory cortex is fundamental to such investigations. Several studies have shown that Correspondence: C. Price, MRC Cyclotron Unit, Hammersmith Hospital, Du Cane Road, London W12 OHS, UK.
hearing words or word-like sounds activates posterior temporal regions [14]. The present study investigated the relationship between the rate of presentation of heard words and rCBF increases in the following way. Six right-handed, English speaking, normal, male volunteers, aged 24-49 years, were studied. Each subject gave informed consent to have 6 consecutive measurements of rCBF, using a C1502 inhalation technique [7, 8] and a Siemens 931-08/12 positron emission tomographic scanner [12]. During each 3.5 min dynamic scan, the subject inhaled CI502 at a concentration of 6 MBq/ml and a flow rate of 500 ml/min through a standard oxygen face mask for a period of 2 rain. Intervals between scans lasted 12-15 rain. Correction for attenuation was made by performing a transmission scan with an exposed 68Ge/ 6SGa external ring source at the beginning of each patient study. Images were reconstructed by filtered back projection (Hanning filter, cut off) giving a transaxial resolution of 8.5 mm full width at half maximum. The reconstructed images contained 128 x 128 pixels, each having a size of 2.05 × 2.05 mm. The behavioral state for each measurement was rest or listening to nouns presented at rates of 10, 30, 50, 70 or 90 words per minute (wpm). The subjects were instructed
180 Left superior temporal gyrus (middle part)
74
}t
}
'°'1 ~s
Left Primary
auditory
} a
cortex
b
c
d
e
f
Right Primary auditory
72S
a8 s
temporal gyrus {middle part)
"'
t
690 67O
{ }
'°~
Right superior
cortex
t
}
66~
630 a
Left posterior
superior
temporal
5c
gyrus
b
c
d
Right posterior
e
f
superior
temporal
gyrus
4S s
tt
4as 42s 41 5 a
b
c
~
e
l
}
4o s a
b
c
d
e
f
Fig. 1. Each graph represents adjusted rCBF plotted against task (a = rest, b = 10 wpm, c = 30 wpm, d = 50 wpm, e = 70 wpm, f= 90 wpm) for all subjects in 6 different regions within the auditory cortex. The coordinates of these regions were selected from the atlas of Talairach and Tournoux [13]. For the middle parts of the superior temporal gyri x = _+50,y -- -22, z - 4 mm; R 2 = 0.92 and 0.96, respectively. For the primary auditory cortices x - + 5 0 , y - - 2 2 , z = I2 mm: R 2 - 0.99 and 0.66, respectively. For the posterior parts of the superior temporalgyrix=_+58, y = 3 4 , z - 12 mm; R2 = 0.12 and 0.88, respectively. The linear relationship between increases in rCBF and presentation rate is illustrated in all regions of the auditory cortex with the exception of the left posterior superior temporal gyrus.
to listen to the tapes o f w o r d s with their eyes closed, o r to relax a n d ' e m p t y their m i n d s ' d u r i n g the rest state. Within the different p r e s e n t a t i o n rate conditions, the w o r d sets were m a t c h e d for w o r d frequency, n u m b e r o f syllables a n d r a t e d imageability. The o r d e r o f a c t i v a t i o n c o n d i t i o n s was d e t e r m i n e d by a L a t i n S q u a r e design.
D a t a were a n a l y z e d using statistical p a r a m e t r i c m a p ping ( S P M Software, M R C C y c l o t r o n Unit, U K ) [3]. C a l c u l a t i o n s a n d image m a t r i x m a n i p u l a t i o n s were perf o r m e d in P R O M A T L A B ( M a t h w o r k s Inc., S h e r b o r n , M A , U S A ) . F o r each subject, r C B F t h r o u g h o u t the whole b r a i n region (15 t r a n s a x i a l planes) was c a l c u l a t e d a c c o r d i n g to a previously described p r o t o c o l [8]. In o r d e r to a c c o u n t for i n d i v i d u a l differences in b r a i n size a n d gyral a n a t o m y , the d a t a were r e o r i e n t a t e d parallel to the i n t e r c o m m i s s u r a l line, then n o r m a l i z e d into the stereotactic space o f the a n a t o m i c a l atlas o f T a l a i r a c h and T o u r n o u x [4, 13]. The v o l u m e elements (voxets) o f these n o r m a l i z e d images (2 x 2 x 4 m m in the x, y a n d z d i m e n sions) where then s m o o t h e d with a G a u s s i a n filter 20 m m wide so t h a t voxel values c o r r e s p o n d e d to a weighted m e a n r C B F in a spherical d o m a i n 20 m m in diameter. These s t a n d a r d i s a t i o n processes also p e r m i t the l o c a t i o n o f each voxel to be referenced to the a n a t o m i c a l atlas o f T a l a i r a c h a n d T o u r n o u x [13]. As the study was designed to examine regional changes across activation c o n d i t i o n s , the d a t a were norm a l i z e d for global flow differences by analysis o f covariance ( A N C O V A ) , with m e a s u r e d g l o b a l flow as the covariate, a n d then a v e r a g e d for each c o n d i t i o n across the 6 subjects [2]. This analysis g e n e r a t e d a d j u s t e d voxel m e a n s for each o f the 6 c o n d i t i o n s a n d the associated adjusted e r r o r variance required for their c o m p a r i s o n . The differences between c o n d i t i o n s were assessed by weighting the c o n d i t i o n m e a n s with the a p p r o p r i a t e contrast. This analysis was d o n e for each voxel a n d the resulting set o f t values c o n s t i t u t e d the t-statistical p a r ametric m a p (SPM{ t}). C o m p a r i s o n o f each a u d i t o r y a c t i v a t i o n c o n d i t i o n c o m p a r e d to rest p r o d u c e d highly significant bilateral increases in r C B F involving the m i d d l e and p o s t e r i o r regions o f the s u p e r i o r t e m p o r a l gyri. F o r the slower rates o f p r e s e n t a t i o n , activity in the left p o s t e r i o r s u p e r i o r
Fig. 2. The 3 brain projections viewing the brain from above (transverse), the right (saginal) and the back (coronal) illustrate that the only regions of the brain where rCBF rises linearly with the presentation rate of heard words were the superior and middle temporal gyri (P < 0.05, corrected for non-independent comparisons). This was assessed with statistical parametric mapping by weighting the six condition means with the rate of word presentation.
181
Fig. 3. The 3 brain projections viewing the brain from above (transverse), the right (sagittal) and the back (coronal) illustrate that the only region in the brain where rCBF rose significantly irrespective of presentation rate was the left posterior superior temporal region (P < 0.05, corrected for non-independent comparisons). This was assessed with statistical parametric mapping by identifying the areas of the brain where the rCBF difference between rest and the slowest rate of presentation was significantlygreater than the average rCBF difference between each word presentation rate. The equation applied was [(b-a) - ({f-b}/4)] where a = rest, b = the slowest rate of word presentation and f = the fastest rate of word presentation.
temporal gyrus was m o r e extensive than activity in the h o m o l o g o u s area on the right but this a s y m m e t r y was less a p p a r e n t at faster rates o f presentation. This observation can be accounted for by plotting the m e a n r C B F values for each condition at different regions o f the temporal gyri. Fig. 1 illustrates that in b o t h p r i m a r y auditory cortices and the middle regions o f the superior temporal gyri, r C B F is linearly correlated with presentation rate. In m o r e posterior regions o f the superior temporal gyri, the linear relationship is present on the right but not on the left. In the left posterior superior temporal region, there is an almost u n i f o r m increase in r C B F irrespective o f presentation rate. Fig. 2 displays the regions o f the brain (superior and middle temporal gyri) where r C B F rises linearly with presentation rate. Fig. 3 illustrates that the only region in the brain where r C B F rose significantly irrespective o f presentation rate is in the left posterior superior temporal gyrus (Wernicke's area). Bilateral temporal lobe lesions can result in pure w o r d s o u n d deafness [1] (an inability to decipher heard words while all other language abilities are intact) whereas lesions that include Wernicke's area result in impaired spoken language in addition to a severe deficit in spoken c o m p r e h e n s i o n [11]. The lesion data therefore supports the n o t i o n that early acoustic analysis o f words occurs in b o t h temporal lobes but Wernicke's area is responsible for the further processing o f encoded verbal material. Activation in Wernicke's area has also been d e m o n strated in P E T studies in association with tasks requiring auditory w o r d c o m p r e h e n s i o n [6, 14]. In the present study Wernicke's area was activated even when subjects are not explicitly asked to r e m e m b e r or understand words, a finding consistent with a b u n d a n t psychological evidence that a u d i t o r y w o r d c o m p r e h e n s i o n proceeds automatically [9]. We conclude that input to areas associated with early
acoustic analysis is dependent on the n u m b e r o f auditory stimuli heard per unit time but the o u t p u t f r o m these areas to Wernicke's area results in a response which is independent o f the n u m b e r o f auditory stimuli heard per unit time. Hence, functional specialisation in Wernicke's area is shown on the basis o f signal t r a n s f o r m a t i o n rather than by association with a specific behavioural task. This w o r k was funded by a G r a n t (91-23) f r o m the M c D o n n e l l Pew Program. 1 Auerbach, S.H., Allard, T., Naeser, M., Alexander, M.P. and Albert, M.L., Pure word deafness: analysis of a case with bilateral lesions and a defect at the prephonemic level, Brain, 105 (1982) 271-300. 2 Friston, K.J., Frith, C.D., Liddle, P.F., Dolan, R.J., Lammertsma, A.A. and Frackowiak, R.S.J., The relationship between local and global changes in PET scans, J. Cereb. Blood Flow Metab., 10 (1991) 458466. 3 Friston, K.J., Frith, C.D., Liddle, EF. and Frackowiak, R.S.J., Comparing functional PET images: the assessment of significant change, J. Cereb. Blood Flow Metab., 11 (1991) 690-699. 4 Friston, K.J., Frith, C.D., Liddle, EF. and Frackowiak, R.S.J., Plastic transformation of PET images, J. Comput. Assist. Tomogr., 15 (1991) 634-639. 5 Fox, P.T. and Raichle, M.E., Stimulus rate determines regional brain blood flow in striate cortex, Ann. Neurol., 17 (1985) 303-305. 6 Howard, D., Patterson, K., Wise, R., Brown, W., Friston, K., Weiller, D. and Frackowiak, R., The cortical localisations of the lexicons: PET evidence, Brain, in press. 7 Lammertsma, A.A., Cunningham, M.P., Deiber, J.D., Heather, EM., Bloomfield, RM., Nutt, J., Frackowiak, R.S.J. and Jones, T., Combination of dynamic and integral methods for generating reproducible functional CBF images, J. Cereb. Blood Flow Metab., l0 (1990) 675-686. 8 Lammertsma, A.A., Frackowiak, R.S.J., Hoffman, J.M., Huang, S., Weinberg, I.N., Dahlbom, M., MacDonald, N.S., Hoffman, E.J., Mazziotta, J.C., Heather, J.D., Forse, G.R., Phelps, M.E. and Jones, T., The C1502 build up technique to measure regional cere-
182 bral blood flow and metabolism, J. Cereb. Blood Flow Metab., 9 (1989) 461~170. 9 Marslen-Wilson, W.D. and Tyler, L.K., Against Modularity. In J. Garfield (Ed.), Modularity and Knowledge Representation and Natural Language Understanding, Lawrence Erlbaum, Hillsdale, NJ, 1987. t0 Raichle, M.E., Circulatory and metabolic correlates of brain function in normal humans. In F. Plum (Ed.), Handbook of Physiology, Vol. 5: The Nervous System, Am. Physiol. Soc., Oxford University Press, New York, 1987, 643-674. 11 Seldon, H.L., The anatomy of speech perception: human auditory cortex. In A. Peters and E.G. Jones (Eds.), Cerebral Cortex, Vol. 4: Association and Auditory Cortices, Plenum, New York, 1985.
12 Spinks, T.J., Jones, T., Gilardi, M.C. and Heather, J.D., Physical performance of the latest generation of commercial PET scanner, IEEE (Inst. Electr. Electron. Eng.) Trans. Nucl. Sci., 35 (1988) 721725. 13 Talairach, J. and Tournoux, P., Co-planar Stereotactic Atlas of the Human Brain, Thieme, Stuttgart, 1988. 14 Wise, R.J.S., Chollet, F., Hadar, U., Friston, K., Hoffner, E. and Frackowiak, R.S.J., Distribution of cortical neural networks involved in word comprehension and word retrieval, Brain, 114 (1991) 1803-1817.
