Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print November 22, 2005, 10.1148/radiol.2381041622
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2381041622v1
238/1/256    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Binder, M.
Right arrow Articles by Urbanik, A. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Binder, M.
Right arrow Articles by Urbanik, A. S.
(Radiology 2005;238:256-263.)
© RSNA, 2005

Neuroradiology

Material-dependent Activation in Prefrontal Cortex: Working Memory for Letters and Texture Patterns—Initial Observations1

Marek Binder, PhD and Andrzej S. Urbanik, MD, PhD

1 From the Department of Psychophysiology, Institute of Psychology (M.B.) and Department of Radiology, Collegium Medicum (A.S.U.), Jagiellonian University, 19 Kopernika St, 31-501 Kraków, Poland. Received September 20, 2004; revision requested November 24; revision received February 2, 2005; final version accepted February 28. Address correspondence to A.S.U. (e-mail: aurbanik{at}mp.pl).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Purpose: To prospectively evaluate whether a distinction between verbal and nonverbal short-term memory systems, as predicted with the multicomponent working memory model, is reflected in the material-specific patterns of activation in the prefrontal cortex.

Materials and Methods: Informed written consent was obtained from all participants, and the institutional review board approved the study protocol. Echo-planar MR imaging was performed in 12 healthy subjects (five female and seven male subjects), with a mean age of 23.52 years ± 2.52 (standard deviation) and a range of 20–29 years. A two-back task was used in the verbal and nonverbal versions. In the first version, letters were used as stimuli, and in the second version, the stimuli were abstract texture patterns. Timing parameters for both versions were the same. Statistical analysis of the functional data involved a fixed-effects general linear model. Regions of activation were identified from specific t-statistic contrasts between baseline and active tasks (corrected for whole-brain multiple comparisons).

Results: The following suprathreshold voxels for the verbal condition were observed predominantly in the left hemisphere (middle frontal gyrus, precentral gyrus, middle temporal gyrus, and occipital cortex). Bilateral activations were in inferior frontal gyri, insulae inferior, superior parietal lobules, and cingulate gyri. In the nonverbal condition, suprathreshold voxels were located mostly bilaterally in the following regions: inferior, middle, and medial frontal gyri and inferior parietal lobules. Active regions were also found in the precentral gyrus and precuneate gyrus in the left parietal lobe and the occipital cortex in the right hemisphere.

Conclusion: Results of this study are consistent with the multicomponent model of working memory.

© RSNA, 2005


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
The multicomponent model of working memory (1–3) assumes an executive attentional component, called central executive, which is supported by the two short-term memory buffers, also known as the slave systems. The first of them, visuospatial sketchpad, serves as the temporary memory for the information about objects and locations. The latter, phonologic loop, is responsible for the active maintenance of speech-based information.

Researchers in several behavioral studies on working memory with task-interference procedures demonstrated that the processing of verbal and nonverbal information is relatively independent and does not compete for the common pool of cognitive resources (4,5). This finding suggests that separate component processes account for both types of information.

Neuroimaging studies about the neural correlates of working memory processes have yielded equivocal results in regard to material-related distinctions within the brain structures. The focus of attention was aimed at the prefrontal cortex, since results of neuropsychological studies indicated that this structure plays the key role in working memory processing (6,7).

Within the prefrontal cortex, lateralization effects for words, drawings, and spatial stimuli have been observed (8–12). For the object and verbal stimuli, the left regions revealed a task-related activation, and for the spatial stimuli, the activated regions were grouped predominantly in the right frontal regions. Other groups of researchers, however, did not discover such heterogeneity with the prefrontal cortex (13–15). Material-dependent activations also are observed in other domains of memory processing: episodic memory encoding and retrieval (16–19) and object-concept activation in the semantic memory (20,21).

The aim of this study was to prospectively evaluate whether a distinction between verbal and nonverbal short-term memory systems, predicted with the multicomponent working memory model of Baddeley (1) and Baddeley and Logie (2), is reflected in the material-specific patterns of activation in the prefrontal cortex.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 References
 
Subjects
Twelve subjects (five female patients, seven male patients) volunteered in this study. They were selected from the local university student community. They were healthy, right-handed native Polish speakers, with no history of neurologic or psychiatric episodes. The health status of the subject was evaluated on the basis of findings at interview and at evaluation of their T1-weighted magnetic resonance (MR) images by one author (A.S.U.). Their mean age was 23.52 years (standard deviation, 2.52), and the age range was 20–29 years. Informed written consent was obtained from all participants, and the institutional review board approved the study protocol.

Cognitive Tasks
The cognitive task we used in this study was the n-back task (22). This task requires the subject to view a series of sequentially presented items (letters or patterns) and to decide whether the current item is identical to the one seen n items back. In this study, we decided to use the value of n = 2 so that subjects were required to monitor the identity of the present item and that of the item seen two screens back.

In the verbal condition, subjects viewed series of letters, whereas in the nonverbal condition, the stimuli were abstract texture patterns. Because we used a blocked design in this study, we also introduced a control condition where n = 0. In this condition, subjects were required to decide whether the currently presented item was identical to the test stimulus. Thus, motor reactions were required during experimental and control conditions. The test stimulus was shown to the subjects before each experimental session. For the verbal stimuli, the test stimulus was an uppercase letter X, and for the nonverbal stimuli, it was a special pattern. The sample stimuli for both versions of each experimental task are shown in Figure 1.



View larger version (13K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1a: Examples of stimuli trains used during (a) zero-back task with letters, (b) two-back task with letters, (c) zero-back task with patterns, and (d) two-back task with patterns. Arrows indicate relevant stimuli that required positive responses from subjects; a motor response to each stimulus was required.

 


View larger version (14K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1b: Examples of stimuli trains used during (a) zero-back task with letters, (b) two-back task with letters, (c) zero-back task with patterns, and (d) two-back task with patterns. Arrows indicate relevant stimuli that required positive responses from subjects; a motor response to each stimulus was required.

 


View larger version (47K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1c: Examples of stimuli trains used during (a) zero-back task with letters, (b) two-back task with letters, (c) zero-back task with patterns, and (d) two-back task with patterns. Arrows indicate relevant stimuli that required positive responses from subjects; a motor response to each stimulus was required.

 


View larger version (46K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1d: Examples of stimuli trains used during (a) zero-back task with letters, (b) two-back task with letters, (c) zero-back task with patterns, and (d) two-back task with patterns. Arrows indicate relevant stimuli that required positive responses from subjects; a motor response to each stimulus was required.

 
Each subject underwent two experimental sessions: one with verbal stimuli and another with nonverbal stimuli. The order of those sessions was counterbalanced across subjects. The experimental design consisted of two experimental two-back blocks and three control zero-back blocks, and these blocks were used in an alternating fashion. Each block lasted for 30 seconds, and each session began with the control condition (Fig 2).



View larger version (13K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2: Diagram shows the block design used in the study.

 
Each stimulus was presented for 1 second, and the interstimulus interval was 2 seconds. At the beginning and the end of each control block, an instruction was presented for 2 seconds, and the instruction signaled the start of the control block and the experimental block, respectively. Thus, during control blocks, participants viewed eight items, whereas during experimental blocks, they viewed 10 items. The letters (Arial Bold font; size, 32 points) were selected from a set containing 17 consonants (B, C, D, F, G, H, J, K, L, M, N, P, R, S, T, W, Z). The V and Q consonants were not used, since they are very rarely used in the subjects' native script of Polish.

Stimuli were back-projected from a multimedia projector (LCD Data Projector VPL-SC50; Sony, Tokyo, Japan) on a screen located about 3 m away from the magnet. Presentation of the stimuli was controlled with in-house software.

Subjects responded in the affirmative or the negative to every stimulus by means of dorsiflexion of the left or right toes, respectively (23). Accuracy of the responses was assigned a score by one of the authors (M.B.). To become familiar with the experimental tasks, subjects practiced for 15 minutes immediately before the beginning of the experiment.

Imaging
The study was performed by using a 1.5-T MR imaging system (Signa Horizon; GE Medical Systems, Waukesha, Wis). The functional images were acquired by using a spin-echo echo-planar imaging sequence, which was sensitive to blood oxygenation level–dependent contrast, with the following parameters: 3000/60 (repetition time msec/echo time msec), a 90° flip angle, a 28 x 21 cm field of view, a 96 x 96 matrix, and one signal acquired. During each functional imaging session, 50 sets of 10 contiguous 7-mm-thick transverse MR images were acquired parallel to the anterior commissure–posterior commissure plane. For each subject, the bottom section was located about 14 mm below the anterior commissure–posterior commissure plane. High-spatial-resolution anatomic images were acquired in the same locations as were the functional images by using the spoiled gradient-recalled acquisition in the steady state sequence with the following parameters: 50/6; flip angle, 60°; field of view, 28 x 21; matrix, 256 x 256; and number of signals acquired, two. A standard radiofrequency head coil with foam padding to restrict head motion was used.

Data Analysis
Performance scores for the verbal and nonverbal conditions were compared by using repeated-measures two-way analysis of variance, with factors of difficulty and stimulus type, by one of the authors (M.B.).

Functional data analysis was performed by both authors in consensus, with a personal computer workstation that was assembled from parts obtained separately from a supplier, by using statistical parametric mapping implemented in statistical software (SPM99; Wellcome Department of Cognitive Neurology, London, England) running on mathematic software (Matlab; Mathworks, Sherbon, Mass).

Spatial preprocessing of data consisted of motion correction, spatial normalization, and smoothing. After we discarded the first two images of each time series to allow magnetic saturation effects, the remaining images were realigned and were motion-corrected in three dimensions by using a six-parameter algorithm. Subsequently, all motion-corrected echo-planar images were normalized to a Montreal Neurological Institute brain template (Montreal Neurological Institute, Montreal, Quebec, Canada) by using an automated algorithm implemented in the statistical software. This algorithm is based on voxel signal intensity measurements and minimizes differences between the functional images and the template by using both linear (translations, rotations, zooms, and shears) and nonlinear (series of discrete cosine basis functions) transformations.

During spatial normalization, we used the procedure proposed by Brett et al (24) to minimize negative effects of the echo-planar imaging magnetic susceptibility artifacts in ventral frontal regions on the effectiveness of the normalization algorithm. This method involves cost-function masking, with which selected areas of the brain are excluded from calculation of the image difference so that these areas do not contribute to any bias in the transformations. Following spatial normalization, images were smoothed with an 8-mm full-width-at-half-maximum isotropic Gaussian kernel.

Images were then filtered by using a high-pass filter of 126 seconds, low-pass filtered by using a hemodynamic impulse response function filter, and then globally scaled. All analyses were performed by using the general linear model that was implemented in the statistical package (25) and that comprised condition and subject effects at each voxel. Effects of task difficulty were calculated with t-statistic contrasts. Correction for multiple comparisons was performed according to the theory of Gaussian fields, and this correction provided a corrected P value (P < .01) for the observed height and spatial extent of each activated cluster (26). We reported only results that had surpassed this threshold. We imposed an extent threshold of 6 voxels on the obtained results.

Because statistical mapping software yields results in the stereotactic space of the Montreal Neurological Institute brain template, we converted resulting cluster coordinates into coordinates described in the atlas of Talairach and Tournoux (27) by using the algorithm of Brett (http://www.mrc-cbu.cam.ac.uk/Imaging/Common/mnispace.shtml). We used software available at http://ric.uthsca.edu/projects/talairachdaemon.html (Talairach Daemon; Research Imaging Center, University of Texas Health Science Center, San Antonio, Tex [28]) to obtain Brodmann area numbers and anatomic labels of the activated clusters.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 References
 
Behavioral Data
The mean percentages of the correct answers are shown in Table 1. Two-way repeated-measures analysis of variance revealed only the main effect of difficulty (F = 6.44, df = 1, 11, P < .05). Neither the effect of the stimulus type nor the interaction effect were significant (P = .120 for effect of stimulus type, P = .147 for interaction effect of stimulus type and difficulty factors). The relatively high standard deviation in the patterns for the two-back condition was caused by the poor performance of one subject whose scores were much lower than those of the rest of the subjects (accuracy, 92%).


View this table:
[in this window]
[in a new window]

  		Table 1. Mean Percentages of Correct Responses for All Experimental Conditions

 
Imaging Data
Statistical maps for the t-statistic contrasts we used revealed multiple sites of activation for both conditions. They were localized predominantly within the frontal and parietal regions. The pattern of the activated clusters was complex for both versions of the experimental task and partially overlapped, although there were considerable differences.

In the verbal task, activation was observed mostly in the left hemisphere (Figs 3, 4). Significantly (P < .01) activated areas are shown in Table 2. Peaks of activity located within the middle frontal gyrus (Brodmann area 46) and precentral gyrus (roughly Brodmann area 9/46) were observed only in the left hemisphere.



View larger version (92K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3: Selected transverse MR image sections overlaid with statistical maps show results of stimulation with nonverbal (patterns) and verbal (letters) versions of the working memory task in the dorsolateral prefrontal cortex. When the stimuli were letters, activation showed up in the left hemisphere, whereas memorization of patterns was correlated with bilateral activation within the prefrontal regions. Background images are a template derived from a high-spatial-resolution (1-mm isotropic voxels) low-noise data set that was created by registering 27 images obtained with T1-weighted gradient-echo sequence (18/10; flip angle, 30°) in the same individual in stereotactic space where they were subsampled and averaged for signal intensity (29). Images are shown in radiologic convention (ie, left side of each image corresponds to subject's right side).

 


View larger version (52K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4: Three-dimensional cortical surface maps illustrate the overall cortical distribution of activation in verbal (letters) and nonverbal (patterns) versions of n-back task. Background images are a template derived from a high-spatial-resolution (1-mm isotropic voxels) low-noise data set that was created by registering 27 images obtained with T1-weighted gradient-echo sequence (18/10; flip angle, 30°) in the same individual in stereotactic space where they were subsampled and averaged for signal intensity (29).

 

View this table:
[in this window]
[in a new window]

  		Table 2. Localization of Activation Clusters and Local Maxima in Verbal Condition

 
Furthermore, the maxima of activation were present bilaterally within the frontal lobes in the following regions: inferior frontal gyri and insulae (Brodmann areas 47 and 13) and cingulate gyri (Brodmann area 32). In the parietal lobes, activation was bilateral, with a strong dominance of the left hemisphere. Peaks of activation were identified in superior and inferior parietal lobules (Brodmann areas 40 and 7). The remaining suprathreshold clusters were localized only in the left hemisphere and were found in the middle temporal gyrus (Brodmann area 22) and in the cuneate and lingual gyri in the left occipital lobe. Within the left hemisphere, we also observed a maximum of activation in the putamen.

The pattern of activated regions revealed with the comparison of the zero-back and two-back nonverbal task differed substantially from the one described in the previous section. Suprathreshold clusters in the frontal lobes were located more bilaterally (Figs 3, 4). As in the verbal task, peaks of activation were observed within the inferior frontal gyri (Brodmann areas 47 and 13), as well as in the left precentral gyrus (Brodmann areas 6 and 9) and the medial frontal gyrus (Brodmann area 32), bilaterally. In contrast to the verbal condition, the local maximum was detected within the middle frontal gyrus region (Brodmann area 46) in both hemispheres. Within the parietal areas, we observed the activation peak within the inferior parietal lobule (Brodmann area 40) bilaterally and in the left precuneate gyrus (Brodmann area 19). In contrast to the results with the verbal version of the task, we did not observe suprathreshold voxels in temporal lobes. Occipital peaks of activation were located within the cuneate gyrus (Brodmann area 30) and in the vicinity of the middle occipital gyrus in the right hemisphere. Maxima of activation were also detected in the putamen. A detailed list of the suprathreshold clusters and their localizations is shown in Tables 2 and 3.


View this table:
[in this window]
[in a new window]

  		Table 3. Localization of Activation Clusters and Local Maxima in Nonverbal Condition

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
References
 
Both cognitive tasks we used revealed a complex pattern of brain activation. The most prominently activated regions were located in the frontal and parietal lobes. Less pronounced sites of activation were found in the occipital and temporal lobes and the subcortical structures. Our main finding was the observation of different patterns of activation for the two versions of the n-back task, and this observation appears to support predictions of the working memory model concerning two separate short-term memory buffers.

The results of other research about verbal working memory with the letter n-back task (8,22) are similar to the results obtained in our study. For example, Braver et al (22) found activation of dorsolateral prefrontal, ventrolateral prefrontal, premotor, and posterior parietal cortices when they compared a letter two-back condition with a zero-back condition. Moreover, the intensity of neural activation of those structures was positively correlated with the n parameter manipulation: It showed a linear increase when this parameter changed from zero to three. Some authors (30,31) regard this complex of cortical regions as a neural basis of the verbal working memory. Within this complex, each structure plays a different role. The dorsolateral prefrontal cortex is responsible for the manipulation of the contents of the verbal short-term memory, the ventrolateral cortex is responsible for the active maintenance and selection of the contents, and the posterior parietal cortex seems to serve as the passive phonologic short-term memory.

Results of recent research (32), however, indicate that the latter can serve more complex functions. The authors suggest that, while the dorsal part of the posterior cortex is dedicated to the general executive processes, the ventral part (supramarginal gyrus), which was not activated in our study, is associated with more basic speech processing, and activity in this region is suppressed in the conditions of heightened memory load. Functional anatomy of the regions involved in the verbal short-term memory processes was the subject of the study by Henson et al (33). Although they did not use the n-back task, a similar pattern of results was found. They emphasized the role of the premotor cortex as a structure involved in maintaining the temporal sequence of the memory set items.

For the nonverbal n-back task, the pattern of obtained results was in general similar to the one observed during the verbal version of this task. Just as was observed there, activations were seen in the frontal and parietal cortices. We found loci of activation within the left prefrontal cortex and in the vicinity of the inferior and middle frontal gyri. Also, activation of the frontal operculum was apparent in both tasks.

In comparison with the verbal task, the main difference is the activation found in the right middle frontal gyrus. This region was active only in the condition where abstract patterns were used. There are two possible interpretations of this result. The first refers to the domain specificity of the processing performed in both hemispheres in the prefrontal regions, and the second explains these differences in terms of different cognitive processes involved in both tasks.

The first interpretation emphasizes different representations in the left and right prefrontal regions. According to that interpretation, analytic-based processing (related to the abstract aspects of the memory material) is carried out by the left prefrontal cortex, while the right prefrontal cortex is involved in the image-based storage and processing of the stimuli (9). Support for this concept comes from results of two series of experiments. The first series involved a direct comparison of the brain activation during tasks that were based on object and spatial stimuli (10,11). For the object stimuli, left hemispheric regions (predominantly in the frontal lobes) revealed heightened activation, and for the spatial stimuli, the activated regions were mainly in the right hemisphere. Results of other series of experiments, which employed faces as stimuli (12), revealed that the activation of the left or the right prefrontal cortex during maintenance of the memory of the faces depended on the delay interval between memory set and probe stimuli. As intervals became longer, activation in the right hemisphere diminished, whereas activation in the left hemisphere gradually increased. According to Courtney et al (9), an iconic representation of the memory material tends to decay over longer periods of time, and more stable analytic-based representation performed by the left hemisphere becomes dominant.

Thus, the bilateral activation seen in the dorsolateral prefrontal cortex during performance of the nonverbal task suggests that subjects attempted to code stimuli both in the form of analytic and iconic representations. This view is closer to the view of Baddeley and Logie (2) in regard to working memory, because, with this view, one also assumes different representations for both types of memory material. For both versions of the task, differences in the activation observed in occipitotemporal regions seem to support this interpretation: For the verbal version, activation was seen in the left hemisphere, and for the nonverbal version, it appeared in the right occipital lobe only.

Working memory–related lateralization of the prefrontal cortex activity also is seen in other types of tasks and groups of subjects. For example, with a verbal and a spatial version of the two-back task, Walter et al (34) observed that schizophrenics did not show expected (spatial, right hemisphere; verbal, left hemisphere) lateralized dominance associated with the type of the task; the latter type of dominance associated with the type of task was evident in the control group. Similar results were found in the study of Reuter-Lorenz et al (35) on verbal and spatial working memory in groups of younger and older adults. By using volume-of-interest analyses, they showed that, in the group of younger adults, the verbal task activated frontal regions in the left hemisphere and the spatial task activated regions in the right hemisphere. In the group of older adults, an anterior bilateral activation for both types of memory was observed.

Another interpretation, however, seems plausible for an explanation of the results of this study. According to it, the additional activation of the right dorsolateral prefrontal cortex is related more to the magnitude of the recruited cognitive resources. In some studies, increasing the demands on the monitoring of working memory contents evoked activation of this structure (36,37). In such cases, observed differences in the prefrontal activation could be accounted for by the different structures that are activated in the cognitive processes involved in both tasks. Processes related to manipulation in working memory (updating, reordering) could explain the activation of the left dorsolateral prefrontal cortex in both tasks (14,30,38–40), and the additional activation of the right dorsolateral prefrontal cortex could be evoked by the increased monitoring demands of the nonverbal tasks, where the stimuli were novel, in contrast to letters, which were familiar to all subjects. This does not exclude the first interpretation, since such monitoring could be performed in the image-based code. Probably an extended experimental design in which stimuli with different degrees of novelty are used and in which the possibility for verbalization exists would provide more clues to resolve issues related to the present results. An example of such a design is that in the study of Golby et al (18) on memory encoding, where four sets of stimuli were used: faces, scenes, abstract patterns, and words. Results showed that the degree of lateralization of prefrontal and medial temporal regions depended on the capability to verbalize stimuli.

Our study had limitations. The number of participants did not enable us to perform a random-effects statistical analysis of the functional data. Such an analysis would make it possible to extend our findings to the whole population. Likewise, having only right-handed subjects did not allow comparison of possible differences in visuospatial abilities of left-handed and right-handed subjects.

As we noted previously, use of a broader scope of the stimuli, differing with regard to the capability of verbal encoding, would provide us more information on the relationship between the laterality of activation and the type of stimuli. Moreover, it would allow us to say more about the task difficulty factor and its influence on the obtained pattern of results.

Another important constraint was the imaging range used in this study; because of technical limitations, it did not include the whole brain volume. Although it included areas that, according to the literature, are usually active during working memory tasks, certainly a wider imaging scope would have offered a more detailed view of the neural network engaged in working memory and any differences associated with verbal and nonverbal stimuli.

Finally, we employed an event-related condition that would have enabled us to have two experimental conditions within one session and would not have required the use of control conditions. Moreover, the event-related condition is not abated by the pure insertion assumption, which is associated with blocked designs.

In conclusion, in this study, we used functional MR imaging to explore neural correlates of the processing involved in working memory. We used two versions of the n-back task: one with letters and the other with texture patterns. The obtained results revealed a complex pattern of neural activity, which involved sites located in the frontal, parietal, and occipital cortices. Although the active areas largely overlapped for both versions of the experimental tasks, we found a material-dependent activation in the dorsolateral prefrontal cortex. We regard these results as consistent with the multicomponent working memory model proposed by Baddeley (1,3) and Baddeley and Logie (2).


   ACKNOWLEDGMENTS
 
The authors thank Barbara Sobiecka, MSc Eng, and Justyna Kozub, MSc, for their invaluable assistance during research.


   FOOTNOTES
 
Authors stated no financial relationship to disclose.

Author contributions: Guarantor of integrity of entire study, M.B.; study concepts/study design or data acquisition or data analysis/interpretation, M.B., A.S.U.; manuscript drafting or manuscript revision for important intellectual content, M.B., A.S.U.; approval of final version of submitted manuscript, M.B., A.S.U.; literature research, M.B., A.S.U.; clinical studies, M.B., A.S.U.; statistical analysis, M.B., A.S.U.; and manuscript editing, M.B., A.S.U.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 

  1. Baddeley A. Working memory. Science 1992;255:556–559.[Abstract/Free Full Text]
  2. Baddeley AD, Logie RH. Working memory: the multiple-component model. In: Miyake A, Shah P, eds. Models of working memory: mechanisms of active maintenance and executive control. New York, NY: Cambridge University Press, 1999; 28–61.
  3. Baddeley AD. Is working memory still working? Am Psychol 2001;56:851–864.[Medline]
  4. Duff SC. What's working in working memory: a role for the Central Executive. Scand J Psychol 2000;41:9–16.[CrossRef][Medline]
  5. Cocchini G, Logie RH, Sala SD, MacPherson SE, Baddeley AD. Concurrent performance of two memory tasks: evidence for domain-specific working memory systems. Mem Cognit 2002;30:1086–1095.[Medline]
  6. Gathercole SE. Neuropsychology and working memory: a review. Neuropsychology 1994;8:494–505.[CrossRef]
  7. Smith EE, Jonides J. Working memory in humans: neuropsychological evidence. In: Gazzaniga MS, ed. The cognitive neurosciences. Cambridge, Mass: MIT Press, 1995; 1009–1020.
  8. Smith EE, Jonides J. Working memory: a view from neuroimaging. Cognit Psychol 1997;33:5–42.[CrossRef][Medline]
  9. Courtney SM, Petit L, Haxby JV, Ungerleider LG. The role of prefrontal cortex in working memory: examining the contents of consciousness. Philos Trans R Soc Lond B Biol Sci 1998;353:1819–1828.[CrossRef][Medline]
  10. Baker SC, Frith CD, Frackowiak RS, Dolan RJ. Active representation of shape and spatial location in man. Cereb Cortex 1996;6:612–619.[Abstract/Free Full Text]
  11. Smith EE, Jonides J, Koeppe RA. Dissociating verbal and spatial working memory using PET. Cereb Cortex 1996;6:11–20.[Abstract/Free Full Text]
  12. Haxby JV, Ungerleider LG, Horwitz B, Rapoport SI, Grady CL. Hemispheric differences in neural systems for face working memory: a PET-rCBF study. Hum Brain Mapp 1995;3:68–82.[CrossRef]
  13. Owen AM, Stern CE, Look RB, Tracey I, Rosen BR, Petrides M. Functional organization of spatial and nonspatial working memory processing within the human lateral frontal cortex. Proc Natl Acad Sci U S A 1998;95:7721–7726.[Abstract/Free Full Text]
  14. Postle BR, D'Esposito M. "What"-Then-"Where" in visual working memory: an event-related fMRI study. J Cogn Neurosci 1999;11:585–597.[Abstract/Free Full Text]
  15. Nystrom LE, Braver TS, Sabb FW, Delgado MR, Noll DC, Cohen JD. Working memory for letters, shapes, and locations: fMRI evidence against stimulus-based regional organization in human prefrontal cortex. Neuroimage 2000;11:424–446.[CrossRef][Medline]
  16. Kelley WM, Miezin FM, McDermott KB, et al. Hemispheric specialization in human dorsal frontal cortex and medial temporal lobe for verbal and nonverbal memory encoding. Neuron 1998;20:927–936.[CrossRef][Medline]
  17. Wagner AD, Poldrack RA, Eldridge LL, Desmond JE, Glover GH, Gabrieli JD. Material-specific lateralization of prefrontal activation during episodic encoding and retrieval. Neuroreport 1998;9:3711–3717.[Medline]
  18. Golby AJ, Poldrack RA, Brewer JB, et al. Material-specific lateralization in the medial temporal lobe and prefrontal cortex during memory encoding. Brain 2001;124:1841–1854.[Abstract/Free Full Text]
  19. Habib R, Nyberg L, Tulving E. Hemispheric asymmetries of memory: the HERA model revisited. Trends Cogn Sci 2003;7:241–245.[CrossRef][Medline]
  20. Kraut MA, Kremen S, Moo LR, Segal JB, Calhoun V, Hart J Jr. Object activation in semantic memory from visual multimodal feature input. J Cogn Neurosci 2002;14:37–47.[Abstract/Free Full Text]
  21. Kraut MA, Kremen S, Segal JB, Calhoun V, Moo LR, Hart J Jr. Object activation from features in the semantic system. J Cogn Neurosci 2002;14:24–36.[Abstract/Free Full Text]
  22. Braver TS, Cohen JD, Nystrom LE, Jonides J, Smith EE, Noll DC. A parametric study of prefrontal cortex involvement in human working memory. Neuroimage 1997;5:49–62.[CrossRef][Medline]
  23. D'Esposito M, Detre JA, Alsop DC, Shin RK, Atlas S, Grossman M. The neural basis of the central executive system of working memory. Nature 1995;378:279–281.[CrossRef][Medline]
  24. Brett M, Leff AP, Rorden C, Ashburner J. Spatial normalization of brain images with focal lesions using cost function masking. Neuroimage 2001;14:486–500.[CrossRef][Medline]
  25. Frackowiak RSJ. Human brain function. San Diego, Calif: Academic Press, 1997.
  26. Worsley KJ, Evans AC, Marrett S, Neelin P. Detecting and estimating regions of activation in CBF activation studies in human brain. In: Uemura K, Lassen N, Jones T, Kanno I, eds. Quantification of brain function: tracer kinetics and image analysis in brain PET. Tokyo, Japan: Excerpta Medica, 1993; 535–548.
  27. Talairach J, Tournoux P. Co-planar stereotaxic atlas of the human brain: 3-dimensional proportional system—an approach to cerebral imaging. Stuttgart, Germany: Thieme Medical Publishers, 1988.
  28. Lancaster JL, Woldorff MG, Parsons LM, et al. Automated Talairach atlas labels for functional brain mapping. Hum Brain Mapp 2000;10:120–131.[CrossRef][Medline]
  29. Collins DL, Zijdenbos AP, Kollokian V, et al. Design and construction of a realistic digital brain phantom. IEEE Trans Med Imaging 1998;17:463–468.[CrossRef][Medline]
  30. D'Esposito M, Postle BR, Rypma B. Prefrontal cortical contributions to working memory: evidence from event-related fMRI studies. Exp Brain Res 2000;133:3–11.[CrossRef][Medline]
  31. Fletcher PC, Henson RN. Frontal lobes and human memory: insights from functional neuroimaging. Brain 2001;124:849–881.[Abstract/Free Full Text]
  32. Ravizza SM, Delgado MR, Chein JM, Becker JT, Fiez JA. Functional dissociations within the inferior parietal cortex in verbal working memory. Neuroimage 2004;22:562–573.[CrossRef][Medline]
  33. Henson RN, Burgess N, Frith CD. Recoding, storage, rehearsal and grouping in verbal short-term memory: an fMRI study. Neuropsychologia 2000;38:426–440.[CrossRef][Medline]
  34. Walter H, Wunderlich AP, Blankenhorn M, et al. No hypofrontality, but absence of prefrontal lateralization comparing verbal and spatial working memory in schizophrenia. Schizophr Res 2003;61:175–184.[CrossRef][Medline]
  35. Reuter-Lorenz PA, Jonides J, Smith EE, et al. Age differences in the frontal lateralization of verbal and spatial working memory revealed by PET. J Cogn Neurosci 2000;12:174–187.[Abstract/Free Full Text]
  36. Owen AM, Evans AC, Petrides M. Evidence for a two-stage model of spatial working memory processing within the lateral frontal cortex: a positron emission tomography study. Cereb Cortex 1996;6:31–38.[Abstract/Free Full Text]
  37. Petrides M, Alivisatos B, Frey S. Differential activation of the human orbital, mid-ventrolateral, and mid-dorsolateral prefrontal cortex during the processing of visual stimuli. Proc Natl Acad Sci U S A 2002;99:5649–5654.[Abstract/Free Full Text]
  38. D'Esposito M, Postle BR. The dependence of span and delayed-response performance on prefrontal cortex. Neuropsychologia 1999;37:1303–1315.[CrossRef][Medline]
  39. Postle BR, Berger JS, D'Esposito M. Functional neuroanatomical double dissociation of mnemonic and executive control processes contributing to working memory performance. Proc Natl Acad Sci U S A 1999;96:12959–12964.[Abstract/Free Full Text]
  40. Wagner AD, Maril A, Bjork RA, Schacter DL. Prefrontal contributions to executive control: fMRI evidence for functional distinctions within lateral prefrontal cortex. Neuroimage 2001;14:1337–1347.[CrossRef][Medline]




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2381041622v1
238/1/256    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Binder, M.
Right arrow Articles by Urbanik, A. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Binder, M.
Right arrow Articles by Urbanik, A. S.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE