Imaging cognition II: An empirical review of 275 PET and fMRI studies

Positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) have been extensively used to explore the functional neuroanatomy of cognitive functions. Here we review 275 PET and fMRI studies of attention (sustained, selective, Stroop, orientation, divided), perception (object, face, space/motion, smell), imagery (object, space/motion), language (written/spoken word recognition, spoken/no spoken response), working memory (verbal/numeric, object, spatial, problem solving), semantic memory retrieval (categorization, generation), episodic memory encoding (verbal, object, spatial), episodic memory retrieval (verbal, nonverbal, success, effort, mode, context), priming (perceptual, conceptual), and procedural memory (conditioning, motor, and nonmotor skill learning). To identify consistent activation patterns associated with these cognitive operations, data from 412 contrasts were summarized at the level of cortical Brodmann's areas, insula, thalamus, medial-temporal lobe (including hippocampus), basal ganglia, and cerebellum. For perception and imagery, activation patterns included primary and secondary regions in the dorsal and ventral pathways. For attention and working memory, activations were usually found in prefrontal and parietal regions. For language and semantic memory retrieval, typical regions included left prefrontal and temporal regions. For episodic memory encoding, consistently activated regions included left prefrontal and medial temporal regions. For episodic memory retrieval, activation patterns included prefrontal, medial temporal, and posterior midline regions. For priming, deactivations in prefrontal (conceptual) or extrastriate (perceptual) regions were consistently seen. For procedural memory, activations were found in motor as well as in non-motor brain areas. Analysis of regional activations across cognitive domains suggested that several brain regions, including the cerebellum, are engaged by a variety of cognitive challenges. These observations are discussed in relation to functional specialization as well as functional integration.     

Neural correlates of confidence during item recognition and source memory retrieval: evidence for both dual-process and strength memory theories

Although the medial-temporal lobes (MTL), PFC, and parietal cortex are considered primary nodes in the episodic memory network, there is much debate regarding the contributions of MTL, PFC, and parietal subregions to recollection versus familiarity (dual-process theory) and the feasibility of accounts on the basis of a single memory strength process (strength theory). To investigate these issues, the current fMRI study measured activity during retrieval of memories that differed quantitatively in terms of strength (high vs. low-confidence trials) and qualitatively in terms of recollection versus familiarity (source vs. item memory tasks). Support for each theory varied depending on which node of the episodic memory network was considered. Results from MTL best fit a dual-process account, as a dissociation was found between a right hippocampal region showing high-confidence activity during the source memory task and bilateral rhinal regions showing high-confidence activity during the item memory task. Within PFC, several left-lateralized regions showed greater activity for source than item memory, consistent with recollective orienting, whereas a right-lateralized ventrolateral area showed low-confidence activity in both tasks, consistent with monitoring processes. Parietal findings were generally consistent with strength theory, with dorsal areas showing low-confidence activity and ventral areas showing high-confidence activity in both tasks. This dissociation fits with an attentional account of parietal functions during episodic retrieval. The results suggest that both dual-process and strength theories are partly correct, highlighting the need for an integrated model that links to more general cognitive theories to account for observed neural activity during episodic memory retrieval.     

Mapping episodic memory

This paper presents an analysis of brain regions generally associated with a frequently used episodic memory task; visual word recognition. The results from five positron emission tomography studies of regional cerebral blood flow, involving a total of nine pairwise comparisons of brain activity related to episodic retrieval and to performance on non-episodic reference tasks, were considered. Across studies, increased activity has been observed in the right anterior and posterior prefrontal, anterior cingulate, bilateral parietal, and cerebellar regions. Decreased activity has been found in bilateral temporal and left midfrontal regions. Comparison of this activation pattern with those of other memory tasks, episodic and semantic, indicate that the right anterior and posterior prefrontal regions guide processes selectively demanded by episodic memory retrieval. There is suggestive evidence from subtraction analyses that these prefrontal regions are activated by different processing components, and analyses of functional connectivity provide further support for functional differentiation. These analyses also point to a critical role of medial-temporal brain regions in episodic retrieval. Taken together, these results show that episodic memory retrieval is mediated by an extensive set of brain regions, some of which seem to be specifically engaged by episodic remembering.     

Effects of incidental and intentional feature binding on recognition: a behavioural and PET activation study.

Using Positron Emission Tomography (PET), we investigated cerebral regions associated with the episodic recognition of words alone and words bound to contextual colours. Two modes of colour encoding were tested: incidental and intentional word-to-colour binding. Word-only recognition was associated with brain activation in a lexico-semantic left middle temporal region and in the cerebellum following an incidental colour encoding, and with brain activation in the left posterior middle frontal gyrus, right anterior cingulate and right inferior frontal gyrus following an intentional encoding. Recognition of bound features was associated with activation in left prefrontal and superior parietal regions following an incidental colour encoding, and with preferential right prefrontal cortex activation following an intentional colour encoding. Our results are in line with the hypothesis of a parietal involvement in context processing, and prefrontal areas in monitoring retrieval processes. Our results also support the hypothesis of a 'cortical asymmetry for reflective activity' (CARA).     

Functional asymmetry of human prefrontal cortex in verbal and non-verbal episodic memory as revealed by fMRI

Functional neuroimaging studies have demonstrated preferential involvement of bilateral prefrontal cortex during episodic memory encoding and retrieval. The aim of the present study is to address the question whether left prefrontal model for encoding holds when highly non-verbal material is used, and which region of the brain is critically related to successful retrieval. To do this, seven normal subjects were investigated using functional magnetic resonance imaging (fMRI) during encoding and retrieval of word and checkerboard pattern. Our results revealed that word encoding activated the left prefrontal cortices and right cerebellum, whereas pattern encoding activated the bilateral middle frontal gyrus, superior parietal lobule, premotor area, and occipital visual cortex. Word-specific activation was found in the ventral prefrontal cortices, and pattern-specific activation located in the right dorsal prefrontal cortex. Conjunction analysis during encoding of word and pattern showed that activity in the left dorsal prefrontal cortex and the right cerebellum might relate to common neural network for encoding regardless of the type of material. Finally, the present study demonstrates strong association between the left ventral prefrontal cortex and retrieval success for word. The evidence, that both encoding and retrieval of words activated the left ventral prefrontal cortex, indicates that this area is involved in active and strategic operation of the mnemonic representation. A lack of the right prefrontal activation during retrieval was interpreted as that activity in this region might relate to retrieval effort rather than success.     

Memory orientation and success: separable neurocognitive components underlying episodic recognition

Episodic recognition can be based on recollection of contextual details, on a sense of recent encounter, or some combination of the two. According to several cognitive models, selectively attending to these distinct aspects of memory may require different retrieval orientations and result in different neural responses depending upon whether or not retrieval is successful. Using event-related fMRI, we examined retrieval orientation by having subjects discriminate between two test words in one of two manners. During source recollection, they selected the member of the pair previously associated with a particular encoding task. In contrast, recency judgment required selection of the most recently encountered item of the pair, regardless of how it had been encoded. Furthermore, successful and unsuccessful trials within each retrieval task were contrasted to determine whether retrieval success effects occurred in overlapping or dissimilar neural populations compared to those associated with each retrieval orientation. The results revealed distinct lateral prefrontal and parietal activations that distinguished attempted source recollection from judgments of relative recency; these orientation effects were largely independent of retrieval success. In contrast, medial temporal lobe structures (hippocampus and parahippocampal gyrus) were differentially more active during successful recollection of encoding context, showing similar reduced responses during failed source recollection and judgments of recency. These results indicate that different memory orientations recruit distinct prefrontal and parietal networks and that the recovery of episodic context is associated with the hippocampus and surrounding medial temporal cortices.     

The effects of divided attention on encoding- and retrieval-related brain activity: A PET study of younger and older adults

Divided attention (DA) disrupts episodic encoding, but has little effect on episodic retrieval. Furthermore, normal aging is associated with episodic memory impairments, and when young adults are made to encode information under DA conditions, their memory performance is reduced and resembles that of old adults working under full attention (FA) conditions. Together, these results suggest a common neurocognitive mechanism by which aging and DA during encoding disrupt memory performance. In the current study, we used PET to investigate younger and older adults' brain activity during encoding and retrieval under FA and DA conditions. In FA conditions, the old adults showed reduced activity in prefrontal regions that younger adults activated preferentially during encoding or retrieval, as well as increased activity in prefrontal regions young adults did not activate. These results indicate that prefrontal functional specificity of episodic memory is reduced by aging. During encoding, DA reduced memory performance, and reduced brain activity in left-prefrontal and medial-temporal lobe regions for both age groups, indicating that DA during encoding interferes with encoding processes that lead to better memory performance. During retrieval, memory performance and retrieval-related brain activity were relatively immune to DA for both age groups, suggesting that DA during retrieval does not interfere with the brain systems necessary for successful retrieval. Finally, left inferior prefrontal activity was reduced similarly by aging and by DA during encoding, suggesting that the behavioral correspondence between these effects is the result of a reduced ability to engage in elaborate encoding operations.     

Age-related changes in parietal lobe activation during an episodic memory retrieval task

The crucial role of lateral parietal regions in episodic memory has been confirmed in previous studies. While aging has an influence on retrieval of episodic memory, it remains to be examined how the involvement of lateral parietal regions in episodic memory changes with age. We investigated episodic memory retrieval in two age groups, using faces as stimuli and retrieval success as a measure of episodic memory. Young and elderly participants showed activation within a similar network, including lateral and medial parietal as well as prefrontal regions, but elderly showed a higher level of brain activation regardless of condition. Furthermore, we examined functional connectivity in the two age groups and found a more extensive network in the young group, including correlations of parietal and prefrontal regions. In the elderly, the overall stronger activation related to memory performance may indicate a compensatory process for a less extensive functional network.     

Neural bases of learning and memory: functional neuroimaging evidence

Positron emission tomography and functional magnetic resonance imaging studies have identified brain regions associated with different forms of memory. Working memory has been associated primarily with the bilateral prefrontal and parietal regions; semantic memory with the left prefrontal and temporal regions; episodic memory encoding with the left prefrontal and medial temporal regions; episodic memory retrieval with the right prefrontal, posterior midline and medial temporal regions; and skill learning with the motor, parietal, and subcortical regions. Recent studies have provided higher specificity, by dissociating the neural correlates of different subcomponents of complex memory tasks, and the cognitive roles of different subregions of larger brain areas.     

Dissociation of frontal and medial temporal lobe activity in maintenance and binding of sequentially presented paired associates

Substructures of the prefrontal cortex (PFC) and the medial-temporal lobe are critical for associating objects presented over time. Previous studies showing frontal and medial-temporal involvement in associative encoding have not addressed the response specificity of these regions to different aspects of the task, which include instructions to associate and binding of stimuli. This study used a novel paradigm to temporally separate these two components of the task by sequential presentation of individual images with or without associative instruction; fMRI was used to investigate the temporal involvement of the PFC and the parahippocampal cortex in encoding each component. Although both regions showed an enhanced response to the second stimulus of a pair, only the PFC had increased activation during the delay preceding a stimulus when associative instruction was given. These findings present new evidence that prefrontal and medial-temporal regions provide distinct temporal contributions during associative memory formation.     

Encoding and retrieving faces and places: Distinguishing process- and stimulus-specific differences in brain activity

Among the most fundamental issues in cognitive neuroscience is how the brain may be organized into process-specific and stimulus-specific regions. In the episodic memory domain, most functional neuroimaging studies have focused on the former dimension, typically investigating the neural correlates of various memory processes. Thus, there is little information about what role stimulus-specific brain regions play in successful memory processes. To address this issue, the present event-related fMRI study used a factorial design to focus on the role of stimulus-specific brain regions, such as the fusiform face area (FFA) and parahippocampal place area (PPA) in successful encoding and retrieval processes. Searching within regions sensitive to faces or places, we identified areas similarly involved in encoding and retrieval, as well as areas differentially involved in encoding or retrieval. Finally, we isolated regions associated with successful memory, regardless of stimulus and process type. There were three main findings. Within face sensitive regions, anterior medial PFC and right FFA displayed equivalent encoding and retrieval success processes whereas left FFA was associated with successful encoding rather than retrieval. Within place sensitive regions, left PPA displayed equivalent encoding and retrieval success processes whereas right PPA was associated with successful encoding rather than retrieval. Finally, medial temporal and prefrontal regions were associated with general memory success, regardless of stimulus or process type. Taken together, our results clarify the contribution of different brain regions to stimulus- and process-specific episodic memory mechanisms.     

Neural correlates of memory for items and for associations: an event-related functional magnetic resonance imaging study

Although results from cognitive psychology, neuropsychology, and behavioral neuroscience clearly suggest that item and associative information in memory rely on partly different brain regions, little is known concerning the differences and similarities that exist between these two types of information as a function of memory stage (i.e., encoding and retrieval). We used event-related functional magnetic resonance imaging to assess neural correlates of item and associative encoding and retrieval of simple images in 18 healthy subjects. During encoding, subjects memorized items and pairs. During retrieval, subjects made item recognition judgments (old vs. new) and associative recognition judgments (intact vs. rearranged). Relative to baseline, item and associative trials activated bilateral medial temporal and prefrontal regions during both encoding and retrieval. Direct contrasts were then performed between item and associative trials for each memory stage. During en- coding, greater prefrontal, hippocampal, and parietal activation was observed for associations, but no significant activation was observed for items at the selected threshold. During recognition, greater activation was observed for associative trials in the left dorsolateral prefrontal cortex and superior parietal lobules bilaterally, whereas item recognition trials showed greater activation of bilateral frontal regions, bilateral anterior medial temporal areas, and the right temporo-parietal junction. Post hoc analyses suggested that the anterior medial temporal activation observed during item recognition was driven mainly by new items, confirming a role for this structure in novelty detection. These results suggest that although some structures such as the medial temporal and prefrontal cortex play a general role in memory, the pattern of activation in these regions can be modulated by the type of information (items or associations) interacting with memory stages.     

An fMRI study of verbal episodic memory encoding in amnestic mild cognitive impairment

Amnestic mild cognitive impairment (aMCI) is a high-risk and often prodromal state for the development of Alzheimer's disease (AD) and is characterised by isolated episodic memory impairment. Functional neuroimaging studies in healthy subjects consistently report left prefrontal cortex (PFC) activation during verbal episodic memory encoding. The PFC activation at encoding is related to semantic processing which enhances memory. The purpose of this study was to ascertain whether impaired verbal episodic memory in aMCI is related to PFC dysfunction. Using functional magnetic resonance imaging (fMRI) we compared 10 aMCI patients with 10 elderly controls during verbal encoding. The encoding task was sensitive to the effects of semantic processing. Subsequent recognition was tested to measure encoding success. Behavioural results revealed impaired recognition and a lower false recognition rate for semantically related distracters (lures) in aMCI, which suggest impaired semantic processing at encoding. Both groups activated left hemispheric PFC, insula, premotor cortex and cerebellum, but group comparisons revealed decreased activation in left ventrolateral PFC in the aMCI group. The magnitude of activation in left ventrolateral PFC during encoding was positively correlated with recognition accuracy in the control group but not in the aMCI group. We propose that verbal episodic memory impairment in aMCI is related to PFC dysfunction which affects semantic processing at encoding.     

Lateralization of prefrontal activity during episodic memory retrieval: evidence for the production-monitoring hypothesis

We propose a new hypothesis concerning the lateralization of prefrontal cortex (PFC) activity during verbal episodic memory retrieval. The hypothesis states that the left PFC is differentially more involved in semantically guided information production than is the right PFC, and that the right PFC is differentially more involved in monitoring and verification than is the left PFC. This "production-monitoring hypothesis" differs from the existing "systematic-heuristic hypothesis," which proposes that the left PFC is primarily involved in systematic retrieval operations, and the right PFC in heuristic retrieval operations. To compare the two hypotheses, we measured PFC activity using positron emission tomography (PET) during the performance of four episodic retrieval tasks: stem cued recall, associative cued recall, context recognition (source memory), and item recognition. Recall tasks emphasized production processes, whereas recognition tasks emphasized monitoring processes. Stem cued recall and context-recognition tasks underscored systematic operations, whereas associative cued recall and item-recognition tasks underscored heuristic operations. Consistent with the production-monitoring hypothesis, the left PFC was more activated for recall than for recognition tasks and the right PFC was more activated for recognition than for recall tasks. Inconsistent with the systematic-heuristic hypothesis, the left PFC was more activated for heuristic than for systematic tasks and the right PFC showed the converse result. Additionally, the study yielded activation differences outside the PFC. In agreement with a previous recall/recognition PET study, anterior cingulate, cerebellar, and striatal regions were more activated for recall than for recognition tasks, and the converse occurred for posterior parietal regions. A right medial temporal lobe region was more activated for stem cued recall and context recognition than for associative cued recall and item recognition, possibly reflecting perceptual integration. In sum, the results provide evidence for the production-monitoring hypothesis and clarify the role of different brain regions typically activated in PET and functional magnetic resonance imaging (fMRI) studies of episodic retrieval.     

Brain regions associated with episodic retrieval in normal aging and Alzheimer's disease.

OBJECTIVE: To examine patterns of brain activation during verbal episodic retrieval in normal elderly subjects and patients in an early phase of AD. BACKGROUND: It is established that 1) a profound episodic memory impairment is a cardinal symptom of AD; and 2) some of the earliest brain changes in this disease occur in regions critical to episodic memory, such as the hippocampus and neighboring regions. Yet, it remains largely unknown whether the episodic memory deficit seen in AD is paralleled by concomitant alterations in brain activity during actual task performance in these or other brain areas. METHODS: Using PET, blood flow was assessed in normal elderly subjects and patients with early AD during two retrieval conditions involving completion of word stems: baseline and cued recall. RESULTS: The patients with AD showed a marked performance deficit in cued recall, although the two groups were indistinguishable in the baseline task condition. Both groups showed bilateral activity in orbital and dorsolateral prefrontal cortex, left precuneus, and right cerebellum, as well as decreased activity in distinct left temporal regions during cued recall. The normal elderly alone activated the left parietal cortex and the left hippocampal formation during episodic retrieval. By contrast, AD-related increases in activity during cued recall were observed in the left orbital prefrontal cortex and left cerebellum. CONCLUSIONS: The similar patterns of activations in the two groups suggest that a large distributed network involved in episodic memory retrieval functions relatively normally in early AD. Those retrieval activations seen in the normal elderly, as opposed to the patients, may reflect AD-related failures in semantic processing and successful recollection of the target information, respectively. Finally, the AD-related increases in activity were interpreted in terms of compensatory reactions to the difficulties in performing the episodic memory task.     

Regional brain activations predicting subsequent memory success: an event-related fMRI study of the influence of encoding tasks

We determined the brain regions that were differentially sensitive to two, randomly inter-mixed tasks: Deep Encoding, in which subjects processed items according to their meaning (is the word pleasant or unpleasant?) and Shallow Encoding, in which items were processed according to two underlined letters in the word (are the letters in alphabetical order?). The former task was associated with activations in a set of brain regions including left lateral prefrontal cortex (PFC) and left medial temporal cortex. The latter showed relatively greater activation in right PFC. Both findings are consistent with predictions made on the basis of previous functional neuroimaging work. Following scanning, each subject underwent a recognition memory task. The results of these provided the basis for a further sub-division of encoding events, according to whether they were predictive of subsequent recognition success or not. Unsurprisingly, recognition performance was greater for words that had been deeply encoded. For both encoding conditions, words that were subsequently recognised were associated with greater activation in a sub-set of regions identified by the deep versus shallow contrast. These included left PFC and medial temporal regions. In left PFC this performance-predicting activation was significantly greater for the deep encoding condition. Our results support previous studies suggesting a role for left PFC and medial temporal cortex in episodic memory encoding. They provide more evidence, too, for a less consistent finding: the interaction between the encoding task and the success of subsequent recognition.     

Hemispheric encoding asymmetry is more apparent than real

Previous neuroimaging studies have claimed a left hemisphere specialization for episodic "encoding" and a right hemisphere specialization for episodic "retrieval." Yet studies of split-brain patients indicate relatively minor memory impairment after disconnection of the two hemispheres. This suggests that both hemispheres are capable of encoding and retrieval. In the present experiment, we examined the possible limits on encoding capacity of each hemisphere by manipulating the "depth" of processing during the encoding of unfamiliar faces and familiar words in the left and right hemispheres of two split-brain patients. Results showed that only the left hemisphere benefited from deeper (more elaborate) encoding of familiar words, and only the right hemisphere benefited from deeper encoding of unfamiliar faces. Our findings are consistent with the view that hemispheric asymmetries in episodic encoding are related to hemisphere-specific processing of particular stimuli. Convergent with recent neuroimaging studies, these results with split-brain patients also suggest that these hemispheric differences are not due to unique specializations in each half brain for encoding memories, but rather, are due to preferential recruitment of the synaptically closer prefrontal cortex to posterior regions processing material-specific information.     

Prefrontal involvement in source memory: an electrophysiological investigation of accounts concerning confidence and accuracy

Although a prefrontal involvement in the memory domain is well-documented, the specific functions the frontal lobes have in episodic memory are still unclear. This study aimed to disentangle theoretical accounts of prefrontal involvement concerning objective characteristics of the retrieval (i.e., accuracy) and accounts based on subjective features (i.e., confidence). Event-related potentials (ERPs) were recorded during the test phase of a source memory task in two experiments. The task was to retrieve the word and the voice of the speaker at study (experiment 1) or the voice of the speaker together with confidence ratings about the source judgment (experiment 2). ERPs in both experiments were not modulated by the success of the voice retrieval, discarding accounts linked to the retrieval success. A right-more-than-left late prefrontal positivity was found in both experiments. Moreover, in experiment 2, waves were more positive for low- than for high-confidence responses. This pattern was observed earlier over lateral parietal scalp regions and later, and more sustained in time, over anterior prefrontal regions. The dissociable effects found within the prefrontal scalp regions, specifically along the anterior-posterior and right-left dimensions, are interpreted as markers of qualitatively different monitoring processes.     

Prefrontal and parietal cortex in human episodic memory: an interference study by repetitive transcranial magnetic stimulation

Neuroimaging findings, including repetitive transcranial magnetic stimulation (rTMS) interference, point to an engagement of prefrontal cortex (PFC) in learning and memory. Whether parietal cortex (PC) activity is causally linked to successful episodic encoding and retrieval is still uncertain. We compared the effects of event-related active or sham rTMS (a rapid-rate train coincident to the very first phases of memoranda presentation) to the left or right intraparietal sulcus, during a standardized episodic memory task of visual scenes, with those obtained in a fully matched sample of subjects who received rTMS on left or right dorsolateral PFC during the same task. In these subjects, specific hemispheric effects of rTMS included interference with encoding after left stimulation and disruption of retrieval after right stimulation. The interference of PC-rTMS on encoding/retrieval performance was negligible, lacking specificity even when higher intensities of stimulation were applied. However, right PC-rTMS of the same intensity lengthened reaction times in the context of a purely attentive visuospatial task. These results suggest that the activity of intraparietal sulci shown in several functional magnetic resonance studies on memory, unlike that of the dorsolateral PFC, is not causally engaged to a useful degree in memory encoding and retrieval of visual scenes. The parietal activations accompanying the memorization processes could reflect the engagement of a widespread brain attentional network, in which interference on a single 'node' is insufficient for an overt disruption of memory performance.     

Further dissociating the processes involved in recognition memory: an FMRI study

Based on an event-related potential study by Rugg et al. [Dissociation of the neural correlates of implicit and explicit memory. Nature, 392, 595-598, 1998], we attempted to isolate the hemodynamic correlates of recollection, familiarity, and implicit memory within a single verbal recognition memory task using event-related fMRI. Words were randomly cued for either deep or shallow processing, and then intermixed with new words for yes/no recognition. The number of studied words was such that, whereas most were recognized ("hits"), an appreciable number of shallow-studied words were not ("misses"). Comparison of deep hits versus shallow hits at test revealed activations in regions including the left inferior parietal gyrus. Comparison of shallow hits versus shallow misses revealed activations in regions including the bilateral intraparietal sulci, the left posterior middle frontal gyrus, and the left frontopolar cortex. Comparison of hits versus correct rejections revealed a relative deactivation in an anterior left medial-temporal region (most likely the perirhinal cortex). Comparison of shallow misses versus correct rejections did not reveal response decreases in any regions expected on the basis of previous imaging studies of priming. Given these and previous data, we associate the left inferior parietal activation with recollection, the left anterior medial-temporal deactivation with familiarity, and the intraparietal and prefrontal responses with target detection. The absence of differences between shallow misses and correct rejections means that the hemodynamic correlates of implicit memory remain unclear.