| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Laboratory of Adaptive Systems, National Institute of Neurological Disorders and Stroke; 2 Rotman Research Institute of Baycrest Centre, University of Toronto, Toronto, Ontario M6A 2E1, Canada; 3 Section on Geriatric Psychiatry, Laboratory of Clinical Science, National Institute of Mental Health; and 4 Positron Emission Tomography Department, National Institutes of Health, Bethesda, Maryland 20892
 |
ABSTRACT |
|---|
 Abstract
 Introduction
Methods
Results
Discussion
References
|
|---|
Schreurs, Bernard G., Anthony R. McIntosh, Marcel Bahro, Peter Herscovitch, Trey Sunderland, and Susan E. Molchan. Lateralization and behavioral correlation of changes in regional cerebral blood flow with classical conditioning of the human eyeblink response. J. Neurophysiol. 77: 2153-2163, 1997. Laterality of changes in regional cerebral blood flow (rCBF) during classical conditioning of the human eyeblink response was studied and changes in rCBF were correlated with conditioned responses. In 10 normal volunteers, rCBF was mapped with positron emission tomography and H215O during pairings of a binaural tone conditioned stimulus and an air puff unconditioned stimulus to the left eye. Control conditions consisted of explicitly unpaired presentations of the tone and air puff before (control) and after (extinction) pairings. During pairings, rCBF increased significantly in right primary auditory cortex (contralateral to air puff) and decreased significantly in left and right cerebellar cortex. There were also increases in rCBF in right auditory association cortex and left temporoccipital cortex. Decreases in rCBF were noted bilaterally in the temporal poles and in the left prefrontal cortex. Positive correlations between changes in rCBF and percent conditioned responses were located in middle cerebellum, right superior temporal cortex, left dorsal premotor cortex, right middle cingulate, and right superior temporal cortex. There were negative correlations in left inferior prefrontal cortex, left middle prefrontal cortex, and right inferior parietal cortex. The data replicate our previous findings of lateralized changes in rCBF following presentations of a binaural tone and air puff to the right eye and indicate that there are pairing-specific changes in primary auditory cortex and cerebellum that are not unique to the left or right hemisphere but are a function of the side of training. The commonalities as well as differences in regional involvement in our present and previous experiment as well as in other eyeblink studies illustrate the advantage of functional neuroimaging to quantify different strategies used by the brain to perform seemingly similar functions. Indeed, the data support the notion that learning-related changes can be detected in a number of specific, but not necessarily invariant, brain regions, and that the involvement of any one region is dependent on the characteristics of the particular learning situation.
The search for neural substrates of learning and memory in model systems has identified a number of sites of plasticity. The use of well-controlled behavioral paradigms, such as classical conditioning, has permitted this search to be focused on a relatively small number of areas (e.g., Thompson 1986 Subjects
The subjects were 10 normal right-handed female volunteers (age 24.5 ± 0.85 yr, mean ± SE, range 21-30 yr) who participated after giving informed consent. Female subjects were used because the experimental equipment allowed little clearance from the scanner; the larger skulls of many males would not allow them and the equipment to pass into the scanner (Molchan et al. 1994 PET scans
PET scans were carried out with a Scanditronix PC2048-15B scanner (Scanditronix) which provides 15 transverse slices through the brain spaced 6.5 mm apart (center to center), with transverse resolution of 6.9 mm (full width at half maximum) and axial resolution of 5-6 mm. Emission scans were obtained after a bolus intravenous injection of 40 mCi (1 Ci = 37 GBq) of H215O (half-life 123 s). Images were acquired over 60 s, starting when the bolus of radio tracer arrived in the head. Because of the near-linear relationship between rCBF and tissue counts accumulated over a brief scan period, the acquired images reflected relative changes in rCBF during different scan states. Six scans were conducted, separated by 12 min to allow for radioactive decay of the H215O to <2% peak levels, except for the second and third scans, which were 19 min apart to allow sufficient time for conditioning to occur.
Eyeblink conditioning
Eyeblinks were detected with the use of a low-torque, rotary potentiometer (Litton) coupled to the upper eyelid. The potentiometer, together with a flexible tube that was positioned ~10 mm from the left cornea (i.e., eye opposite that used by Molchan et al. 1994 Image analysis
POSTPROCESSING.
Images from each subject were linearly interpolated from the original 15 slices to 43, and then images from the second through sixth scan were registered to the first scan with the use of algorithms developed by Woods et al. (1992) STATISTICAL ANALYSIS.
The characterization of learning-related changes in rCBF was performed with the use of an analytic approach involving a univariate test (Molchan et al. 1994 SPM.
A repeated-measures analysis of variance procedure was performed on a voxel-by-voxel basis with an analysis of covariance adjustment for variations in global blood flow (Friston et al. 1991a PLS ANALYSIS.
PLS analysis operates on the interrelation among two or more "blocks" of matrices and seeks to obtain a new set of variables that optimally relates the blocks with the use of the fewest dimensions. In the present application, one block contains matrices of the functional PET images for all subjects in all experimental conditions, a second block consists of contrasts coding the experimental design, and a third block contains the percent CRs or CR latencies. Analyzing the second and third blocks against the first, PLS can carry out image-wide PET activation analysis and extract brain-behavior covariances and correlations. Within the space of image descriptors, the findings of PLS are carried by singular images, computed multivariate optima somewhat related to the previously introduced eigenimages of Friston et al. (1993) STATISTICAL SIGNIFICANCE.
Averaging the scores within a task and comparing these averages between tasks provides an indication of whether a particular spatial pattern significantly relates to the contrasts. A descriptive index of this discrimination can be obtained by the use of something like a one-way analysis of variance; however, the scores may be biased and therefore conventional parametric statistics may not be appropriate. The inferential interpretation of the task discrimination by the scores obtained by PLS was assessed with the use of a permutation test (Edgington 1980 BRAIN-BEHAVIOR RELATIONS.
In many neuroimaging studies, behavioral measures are obtained when subjects are engaged in the experiment. Aside from ensuring that subject performance is within expectation, behavioral measures can also provide important information relating brain activity with observed behavior. A cross-correlation of subject behavioral measures and brain activity within task is straightforward. However, with the use of PLS, not only can this correlation be examined, but changes in brain-behavior correlations across tasks can be assessed also. Of particular interest in this analysis was whether similar brain regions were identified as showing a relation to behavior as those regions identified in the analysis of changes in rCBF. The analysis differs from that concerned with experimental design. Rather than the covariance of brain and design contrasts, the cross-correlation matrix contains the relation between the behavioral measure and brain within each condition of interest (e.g., in the case of 2 scans, the matrix would have 2 cross-correlation vectors). The resulting singular images thus indicate whether there are common patterns of brain-behavior relations between the conditions by extracting singular images that can potentially represent commonalities or differences where they exist. This form of PLS analysis is more similar to applications in teratology (Streissguth et al. 1993).
Behavior
Figure 1 shows sample responses to the tone and air puff during the course of the experiment. From top to bottom, the figure shows an unconditioned response to the tone (alpha response); an unconditioned response to the air puff; a CR during pairing of the tone and air puff; and a combination of alpha response and CR during pairing.
SPM image analysis of classical conditioning
Results from the comparison of averaged rCBF during the paired and unpaired control scans are summarized in Table 1 and the corresponding difference images are depicted in Fig. 3, top. Each of these areas showed not only a significant difference in rCBF between the paired and unpaired control phase but also a significant quadratic trend across the three phases of the experiment. Consequently, the changes in rCBF in these areas followed the associative relationship between tone and air puff. Inspection of the figure and table shows both increases and decreases in rCBF as a function of conditioning. There were significant increases in the primary auditory [Brodmann area (BA) 42; 56,
SPM image analysis of extinction
Results from the comparison of averaged rCBF during the paired and unpaired extinction scans are summarized in Table 2 and the corresponding difference images are depicted in Fig. 3, bottom. Inspection of the figure shows both increases and decreases in rCBF as a function of unpaired extinction. There was an increase in rCBF that also contained a quadratic trend in the right cerebellar cortex (16,
PLS analysis
Helmert orthogonal contrasts were used for the design matrix in the PLS analysis. This simply codes for the total degrees of freedom for the experimental effects (total number of scans minus 1), rather than performing pairwise contrasts as used in SPM. It is important to note that the outcome of the PLS analysis would be identical regardless of the exact contrasts used so long as all the degrees of freedom were represented (McIntosh et al. 1996
Behavioral correlations
The PLS analysis of correlations of the brain-behavior relations was performed for the two paired scans and the two unpaired extinction scans. Analyses of percent CRs and CR latency were each performed separately. In both cases, there were strong similarities in the singular images identified by looking at either percent CRs or latency measures, with some notable exceptions.
The present experiment replicates and extends our previous findings in which classical conditioning of the eyeblink response produced increases in rCBF in the primary auditory cortex opposite the side of air puff delivery and decreases in rCBF in the cerebellar cortex on the same side as the air puff (Molchan et al. 1994
INTRODUCTION
Abstract
Introduction
Methods
Results
Discussion
References
; for review see Schreurs 1989). For example, studies of classical conditioning of the rabbit nictitating membrane/eyelid response have focused attention on the cerebellum as a possible site underlying learning (see Thompson and Krupa 1994). In addition, evidence from clinical studies suggests that the human cerebellum may be crucial to eyeblink conditioning (Daum et al. 1993; Topka et al. 1993). There is also considerable evidence from both human and animal learning literature that primary sensory cortices are important in learning and memory beyond simple stimulus processing. Functional plasticity correlated with learning and memory has been demonstrated at the cellular level in single sensory cortical cells (Recanzone et al. 1992; Weinberger et al. 1990) and in cellular ensembles in subcortical and cortical maps (Gonzalez-Lima 1992; Harvey et al. 1986; Scheich et al. 1992). This plasticity has recently been shown to occur in the human auditory cortex with the use of positron emission tomography (PET) regional cerebral blood flow (rCBF) (Molchan et al. 1994). These data suggest that sensory systems can code, in parallel, the physical and behavioral properties of stimuli. Such findings were anticipated in classical conditioning studies of the marine mollusk, Hermissenda, where a pair of type B photoreceptors has been identified as the site of learning-specific change (e.g., Alkon 1983).
showed that during classical conditioning of the human eyeblink response there was a significant increase in rCBF in the primary auditory cortex and a significant decrease in rCBF in the cerebellum. A striking aspect of those data was the considerable laterality in the changes in blood flow. Specifically, changes in the auditory cortex were strongest on the side opposite to the air puff delivery, whereas cerebellar changes were strongest on the same side as the air puff. Interestingly, the tone was presented binaurally and observations of human eyeblink conditioning of the right eye show that conditioned responses (CRs) are markedly bilateral (e.g., Hilgard and Campbell 1936). In addition to lateralized changes in auditory cortex and cerebellum, Molchan et al. (1994) identified a number of other areas including cingulate cortex, parietal cortex, and prefrontal lobes that are involved in eyeblink conditioning. These data suggest that a form of learning as apparently simple as eyeblink conditioning appears to engage an extensive network of neural systems. Because the involvement of these areas was specific to the acquisition of the CR, the data of Molchan et al. (1994) lend support to the notion that learning and memory may involve the interactions among a number of neural systems and that any region has the potential to play a role depending on the requirements of the task (McIntosh and Gonzalez-Lima 1994b).
). In addition, the present experiment sought to correlate changes in rCBF with behavior (% CRs and latency of CRs) and elaborate on the complexity of changes in rCBF that occur with learning. Areas that showed strong relations to the acquisition of the CR may suggest a special role for those areas in learning the behavior.
METHODS
Abstract
Introduction
Methods
Results
Discussion
References
). Subjects were given instructions that would not reveal the nature of relationship between the tone and air puff during the experiment (neutral instruction set). The study was approved by appropriate institutional review boards at the National Institutes of Health.
.
) to deliver a 100-ms, 2-psi (1 psi = 6.89 kPa) air puff, was attached to the left side of a plastic mask, which served as the head-holding device for the PET scans. The 2-psi air puff served as the unconditioned stimulus (UCS) and was strong enough to elicit an eyeblink reliably without being noxious. Earphones delivered a bilateral 80-dB, 1,000-Hz, 500-ms tone, which served as the conditioned stimulus (CS). The output of the potentiometer was amplified and recorded on a chart recorder (Gould model 2200), operated at a chart speed of 125 mm/s.
. The change was made to provide the same number of tone and air puff presentations in each phase and so ensure greater comparability of the extinction phase with the unpaired control and paired phases. Consequently, in the present experiment, the number of tone and air puff presentations was the same for each of the three phases. During scans the intertrial intervals for paired trials averaged 15 s and the intertrial interval for unpaired trials averaged 8 s. Paired presentations of tone and air puff began after the second unpaired control scan and continued through the first and second paired scans and between the first and second unpaired extinction scans. The paired presentations that occurred between scans were separated by an average intertrial interval of 30 s.
150 ms after tone onset but before air puff onset. On explicitly unpaired extinction trials, a CR was defined as any eyelid closure >0.5 mm that occurred 150 ms after tone onset but before the end of the 1,200-ms observation interval. Responses that occurred <150 ms after tone onset were scored as alpha responses (unconditioned responses to the tone) and not counted as CRs (see for example, Gormezano 1966).
to correct for head movement across the experiment. After registration, subject images were spatially transformed to an averaged PET image that conformed to the standard space (Friston et al. 1989, 1991a). Transformed images were then smoothed with a 20 × 20 × 12 mm Gaussian filter to reduce individual differences in functional anatomy.
) followed by a new, more sensitive multivariate image analysis. Specifically, Statistical Parametric Mapping software (SPM, version 4) was used to assess task-related activity changes pixel by pixel (Molchan et al. 1994). After this a multivariate analysis, known as Partial Least Squares, identified distributed patterns of regional activity that most related the experimental design (PLS) (Bookstein et al. 1996; Streissguth et al. 1993; see McIntosh et al. 1996 for the extension to image analysis). This approach can be seen as using the SPM analysis for regional tests ("is this region significantly activated across tasks?") and the PLS analysis to test distributed systems ("are there systems that as a whole are related to the task?"). Although there will be some overlap in the regions identified by each analysis, the main distinction is that the PLS analysis works at the system level, whereas the SPM analysis works at the regional level. Each of these analyses is explained more fully below.
,b). There were no systematic differences in global cerebral blood flow between tasks [F(5,54) = 2.10, P > 0.05], so the validity of the analysis of covariance adjustment was confirmed. Pairwise comparisons and linear and quadratic trend analyses were conducted across the three different phases of the experiment (significance level of P < 0.001). Linear increases or decreases in rCBF across the phases of the experiment might indicate sensitizing or habituating effects that resulted from stimulus presentation and/or duration in the scanner. In contrast, quadratic trends in rCBF would follow the associative relationship between tone and air puff across the three phases of the experiment (i.e., unrelated, related, and then unrelated again, see Molchan et al. 1994).
. The singular images are spatial patterns of pixel contents representing pixel-by-pixel covariances with task or behavior. The numerical values within the singular images are called saliences.
; McIntosh et al. 1996). For the present application, the test involves randomly reordering the rows of the design matrix and redoing PLS and analysis of variance with the new ordering. At each permutation, the statistic is compared with the obtained statistic for the original data. The obtained statistic is assigned a probability value based on the number of times a statistic from the permuted data exceeds the obtained value. With the use of this method, the significance of the PLS scores can be assessed without relying on the distributional assumptions underlying most conventional parametric statistical methods. The permutation test used in this application was a multiple regression approach to analysis of variance (Pedhazur 1982). The collection of condition contrast vectors (or behavioral measures) was used as independent variables and the scores as dependent variables.
RESULTS
Abstract
Introduction
Methods
Results
Discussion
References
View larger version (8K):
[in a new window]
FIG. 1.
Sample responses to the tone and air puff during the course of the experiment. Top to bottom: an unconditioned response to the tone (alpha response); an unconditioned response to the air puff (UR); a conditioned response (CR) during pairing of the tone and air puff; a combination of alpha response and CR during pairing; and duration and timing of the tone conditioned stimulus (CS) and air puff unconditioned stimulus (UCS). Responses were digitized from chart recordings obtained during the course of the experiment.
View larger version (10K):
[in a new window]
FIG. 2.
Percent CR and alpha responses during each scan (UP1 and UP2, unpaired tone and air puff; P1 and P2, paired tone and air puff; EX1 and EX2, unpaired tone and air puff extinction) and for 9 6-min blocks (B1, B2, . . . B9) of tone and air puff pairings between scans.
; Myers and Gluck 1994). This interpretation is supported by an examination of CR latencies that did not change between pairings and scans, averaging 313 ms during paired training and 312 ms during the paired scans.
8, 8], auditory association (BA 22; 30, 44, 8), and temporoccipital cortices (BA 18; 32, 86, 8). All of these increases, with the exception of the increase in temporoccipital cortex, occurred in the right hemisphere (i.e., contralateral to the side of UCS presentation). In contrast, there were a number of decreases that occurred in the left hemisphere and others that occurred bilaterally. Specifically, there were decreases in the left prefrontal cortex (BA 47; 42, 32, 12) and bilateral decreases in the cerebellar cortex (42, 72, 28 and 16, 72, 28) and the temporal poles (e.g., BA 38; 24, 14, 28).
View this table:
TABLE 1.
Comparison of averaged rCBF during paired scans vs. unpaired control scans
View larger version (K):
[in a new window]
FIG. 3.
Z-transformed difference images for the comparison of the averaged paired minus unpaired control scans (top) and averaged unpaired extinction minus paired scans (bottom). The atlas plates (Talairach and Tornoux 1988 ) show the location (in mm) of the brain section relative to a line connecting the anterior and posterior commissures. Images are in the transverse plane: anterior is at top and left is at left. Only slices in which areas showed both statistically significant differences and quadratic trend are shown. Statistically significant increases in regional cerebral blood flow (rCBF) are graded from red to yellow with red at the threshold of P < 0.001. Statistically significant decreases in rCBF are depicted in blue, which corresponds to a threshold of P < 0.001. The stereotaxic coordinates for the maximum differences are listed in Tables 1 and 2.
20, 36, 28), left anterior cingulate (20, 2, 32), and right posterior cingulate (12, 30, 28). Decreases in rCBF were noted in the left hippocampus (28, 4, 16), medial cerebellum (30, 68, 20), left medial temporoccipital lobe (40, 38, 12), and superior temporal lobe (58, 50, 20).
68, 28). Decreases in rCBF during unpaired extinction that were also featured in a quadratic trend analysis occurred in the left temporoccipital lobe (BA 18; 34, 84, 4) and the right superior and transverse temporal cortices (BA 22; 48, 12, 0, and BA 42; 50, 6, 4, respectively). Finally, areas that showed both a decreased rCBF during unpaired extinction and a linear trend included the right inferior temporal cortex (42, 66, 0), the right anterior cerebellar cortex (46, 32, 24), and the right inferior parietal lobe (40, 30, 32).
View this table:
TABLE 2.
Comparison of averaged rCBF during unpaired extinction scans vs. paired scans
).
View larger version (38K):
[in a new window]
FIG. 4.
Results from Partial Least Squares (PLS) analysis of activation data. The singular images present orthogonal projections plots of a "glass brain" showing positive and negative saliences. Scores for each subject are plotted by condition. The scores have been transformed to standard space. A: pattern that related to a linear "time" effect across the experiment. B: pattern correlated with the change in the associative relation of the CS and UCS and therefore reflecting areas most related to acquisition and extinction of the CR.
12, 22, 40) and orbitofrontal cortices (BA 11; 20, 36, 12, and 16, 40, 12), which followed a linear decrease, and the lateral cerebellum (26, 68, 20) and anterior temporal cortices (BA 20; 40, 18, 24, and 34, 0, 24), which showed linear increases. For the second pattern, the left inferior (BA 47; 48, 36, 4) and middle prefrontal cortices (BA 9; 42, 18, 24) and middle cerebellum (2, 72, 28) all followed a U-shaped function, whereas the left parahippocampal gyrus near the uncus (BA35; 22, 22, 20), bilateral temporoccipital cortices (BA 22/21; 38, 56, 8 and 42, 60, 8), and right auditory cortex (BA 41/42; 42, 12, 4) was negatively related to this U-shaped pattern. The second pattern therefore identified areas that were specifically related to the change in the associative relationship between the CS and UCS, independent of effects accounted for by the first pattern. Moreover, the pattern corresponds very closely to the quadratic trend observed in the SPM analysis as well as identifying the middle prefrontal cortices and the hippocampus. Variation in the location of the maximal regions of change is related to the different sources of error variance in the univariate SPM versus multivariate PLS (see Arndt et al. 1995 for further discussion of this issue).
50, 28), left dorsal premotor cortex (BA 6/9; 10, 30, 24), right middle cingulate (BA 24; 12, 20, 36), and right superior temporal cortex (BA 22; 40, 46, 8). Areas showing the opposite relation to CRs (negative saliences or decreased rCBF with increased CRs) were left inferior prefrontal cortex (BA 45/47; 38, 24, 0), left middle prefrontal cortex (BA 9; 30, 14, 32), and right inferior parietal cortex (BA 40; 30, 26, 24).
DISCUSSION
Abstract
Introduction
Methods
Results
Discussion
References
). A number of others areas associated with aspects of learning and memory were also found to change as a function of classical conditioning of the eyeblink response. In particular, there were increases in rCBF in auditory association cortex and temporoccipital cortex, and decreases in the temporal poles and inferior prefrontal lobe.
dorsal prefrontal area 9 and inferior prefrontal area 45/47. These areas are anatomically distinct and do have different connection patterns (Pandya and Yeterain 1990). These areas may be involved in different parts of the associative process (e.g., execution of CR, identification of CS-UCS relationship). Among many other things, the dorsal prefrontal cortex has been identified as playing some role in associative behaviors. Indeed, it has been demonstrated that patients with frontal lobe damage have difficulty in associative tasks (conditional associative learning, CAL task, Petrides 1985). Although eyeblink conditioning may be somewhat simpler than the CAL task, our data do support an important role for dorsal prefrontal cortex that spans all stages of associative learning. The inferior prefrontal cortex, on the other hand, was identified in our previous experiment (Molchan et al. 1994) and has been anatomically linked to the temporal lobes (Pandya and Yeterain 1990), leading to the supposition that this region may play a key role in the sensory association between CS and UCS. The dorsal prefrontal cortex, by virtue of its strong ties to premotor and motor cortices, may act to coordinate the sensory information with the motor response. In our previous study, the prefrontal activity was right lateralized, ipsilateral to the side of the UCS. Here too the prefrontal involvement was lateralized to the side of UCS presentation.
; Logan and Grafton 1995), which provide significant confirmation of some of our own data as well as identifying some differences. Logan and Grafton (1995) examined glucose metabolism and found increases in a number of regions including the cerebellum, hippocampus, striatum, temporal gyrus, and occipitotemporal fissure. Blaxton et al. (1996) examined rCBF and found increases in striatum and hippocampus, and decreases in cerebellum. Clearly, a number of areas, most notably the cerebellum and hippocampus, were identified to be specifically involved during eyeblink conditioning in all of the imaging studies. Areas such as the striatum, temporal gyrus, and temporoccipital cortex were found to be important in at least two of the studies. Differences in imaging methods (e.g., glucose metabolism vs. rCBF, different fields of view; the Logan and Grafton study positioned the PET camera more ventrally), behavioral paradigms (massed vs. continuous training), and conditioning levels may account for some of the between-study differences in brain areas and differences in the direction of changes observed in the imaging of human eyeblink conditioning. These factors may also combine with sampling different time windows during the acquisition of the CR. It is possible that certain areas are recruited at different times during acquisition and may not be identified depending on what portion of the acquisition curve is sampled. It may be possible to remedy these discrepancies either by obtaining more scans during different stages of acquisition, or to use a method that affords better temporal resolution to follow the acquisition curve continuously (e.g., evoked response potentials, functional magnetic resonance imaging).
. Second, there were several areas identified in the Molchan et al. (1994) study (e.g., striatum and parietal and insular cortices) that did not appear in the present experiment and several areas identified in the present experiment (e.g., hippocampus, temporoccipital cortex, temporal poles) that were not found in the original study. Some of these differences may be explained by the fact that the overall levels of conditioning were somewhat lower in the present study (62.5% CRs) than in our earlier experiment (73.7% CRs). However, all of the areas identified across our two studies have been found to be involved in associative and/or attentional processes (e.g., inferior temporal cortex, Miyashita 1993; occipitotemporal cortex, Walsh and Perrett 1994), so the differences between studies may reflect the phasic involvement of different areas during different stages of response acquisition. For example, all of the areas identified in our present and previous study were also identified in recent imaging studies of human eyeblink conditioning (see above), which examined the brain with the use of a different radiotracer (Logan and Grafton 1995) and at different points during the acquisition of the conditioned eyeblink response (e.g., Blaxton et al. 1996; Logan and Grafton 1995).
; Pardo et al. 1990). Working memory tasks have been used extensively in PET with some discrepant results. However, a recent study by Haxby et al. (1995) found that different areas become active if working memory load is parametrically manipulated, suggesting that some of the specific regions engaged in working memory are not fixed. All of these studies suggest that regional participation in different functions may very much depend on the specific requirements of the task. The commonalities and discrepancies in regional involvement across different experiments concerning similar phenomena should be taken as an advantage of functional neuroimaging studies. The ability to explicitly quantify the different strategies used by the brain to perform seemingly similar functions is not attainable through classic approaches of neuroscientific investigation, such as ablation methods. These methods can characterize the dysfunctions arising from damage, but are not able to identify the distributed system(s) that underlie normal performance.
). This lends support to a growing body of literature suggesting that primary sensory cortices have an important role in associative learning beyond simple stimulus processing (Bakin and Weinberger 1990; Gonzalez-Lima and Scheich 1986; Harvey et al. 1986; Molchan et al. 1994). This point is reinforced by our observation that presentation of a binaural tone again resulted in predominantly unilateral change in auditory cortex. Moreover, the fact that changes in rCBF in the auditory cortex occurred on the side opposite the air puff rather than simply in the left auditory cortex (Molchan et al. 1994) strengthens the idea that sites of learning-specific modification are not fixed at a specific anatomic location but are functionally related to the nature of the stimuli involved and the task required. The decrease in rCBF bilaterally in the cerebellar cortex also replicates our earlier finding and adds further evidence to the involvement of the cerebellar cortex in eyelid conditioning. Indeed, the data lend support to a growing body of imaging literature suggesting that the cerebellum is involved in learning and memory (e.g., Andreasen et al. 1995; Blaxton et al. 1996; Logan and Grafton 1995). Finally, the present findings lend further credence to the notion that normal brain functioning during associative learning involves cooperative interactions among a number of brain areas (McIntosh and Gonzalez-Lima 1994a; Molchan et al. 1994).
| |
FOOTNOTES |
|---|
Address for reprint requests: B. G. Schreurs, Bldg. 36, Rm. B205, NINDS, NIH, Bethesda, MD 20892.
Received 24 April 1996; accepted in final form 3 December 1996.
| |
REFERENCES |
|---|
|
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
This article has been cited by other articles:
|
|
 |
|
  D. T. Cheng, J. F. Disterhoft, J. M. Power, D. A. Ellis, and J. E. Desmond Neural substrates underlying human delay and trace eyeblink conditioning PNAS, June 10, 2008; 105(23): 8108 - 8113. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Miller, C. Weiss, X. Song, G. Iordanescu, J. F. Disterhoft, and A. M. Wyrwicz Functional Magnetic Resonance Imaging of Delay and Trace Eyeblink Conditioning in the Primary Visual Cortex of the Rabbit J. Neurosci., May 7, 2008; 28(19): 4974 - 4981. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. B. Caplan, A. R. McIntosh, and E. De Rosa Two Distinct Functional Networks for Successful Resolution of Proactive Interference Cereb Cortex, July 1, 2007; 17(7): 1650 - 1663. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Kronenbuerger, M. Gerwig, B. Brol, F. Block, and D. Timmann Eyeblink conditioning is impaired in subjects with essential tremor Brain, June 1, 2007; 130(6): 1538 - 1551. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Alain, J. S. Snyder, Y. He, and K. S. Reinke Changes in Auditory Cortex Parallel Rapid Perceptual Learning Cereb Cortex, May 1, 2007; 17(5): 1074 - 1084. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Gottselig, D. Brandeis, G. Hofer-Tinguely, A. A. Borbely, and P. Achermann Human Central Auditory Plasticity Associated With Tone Sequence Learning Learn. Mem., March 1, 2004; 11(2): 162 - 171. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Miller, N.-k. Chen, L. Li, B. Tom, C. Weiss, J. F. Disterhoft, and A. M. Wyrwicz fMRI of the Conscious Rabbit during Unilateral Classical Eyeblink Conditioning Reveals Bilateral Cerebellar Activation J. Neurosci., December 17, 2003; 23(37): 11753 - 11758. [Abstract] [Full Text] |
