Neural correlates of semantic and episodic memory retrieval

Meta-analyzing left hemisphere language areas: phonology, semantics, and sentence processing

The advent of functional neuroimaging has allowed tremendous advances in our understanding of brain-language relationships, in addition to generating substantial empirical data on this subject in the form of thousands of activation peak coordinates reported in a decade of language studies. We performed a large-scale meta-analysis of this literature, aimed at defining the composition of the phonological, semantic, and sentence processing networks in the frontal, temporal, and inferior parietal regions of the left cerebral hemisphere. For each of these language components, activation peaks issued from relevant component-specific contrasts were submitted to a spatial clustering algorithm, which gathered activation peaks on the basis of their relative distance in the MNI space. From a sample of 730 activation peaks extracted from 129 scientific reports selected among 260, we isolated 30 activation clusters, defining the functional fields constituting three distributed networks of frontal and temporal areas and revealing the functional organization of the left hemisphere for language. The functional role of each activation cluster is discussed based on the nature of the tasks in which it was involved. This meta-analysis sheds light on several contemporary issues, notably on the fine-scale functional architecture of the inferior frontal gyrus for phonological and semantic processing, the evidence for an elementary audio-motor loop involved in both comprehension and production of syllables including the primary auditory areas and the motor mouth area, evidence of areas of overlap between phonological and semantic processing, in particular at the location of the selective human voice area that was the seat of partial overlap of the three language components, the evidence of a cortical area in the pars opercularis of the inferior frontal gyrus dedicated to syntactic processing and in the posterior part of the superior temporal gyrus a region selectively activated by sentence and text processing, and the hypothesis that different working memory perception-actions loops are identifiable for the different language components. These results argue for large-scale architecture networks rather than modular organization of language in the left hemisphere.

Automatic Priming of Semantically Related Words Reduces Activity in the Fusiform Gyrus

We used rapid, event-related fMRI to identify the neural systems underlying object semantics. During scanning, subjects silently read rapidly presented word pairs (150 msec, SOA = 250 msec) that were either unrelated in meaning (ankle-carrot), semantically related (fork-cup), or identical (crow-crow). Activity in the left posterior region of the fusiform gyrus and left inferior frontal cortex was modulated by word-pair relationship. Semantically related pairs yielded less activity than unrelated pairs, but greater activity than identical pairs, mirroring the pattern of behavioral facilitation as measured by word reading times. These findings provide strong support for the involvement of these areas in the automatic processing of object meaning. In addition, words referring to animate objects produced greater activity in the lateral region of the fusiform gyri, right superior temporal sulcus, and medial region of the occipital lobe relative to manmade, manipulable objects, whereas words referring to manmade, manipulable objects produced greater activity in the left ventral premotor, left anterior cingulate, and bilateral parietal cortices relative to animate objects. These findings are consistent with the dissociation between these areas based on sensory-and motor-related object properties, providing further evidence that conceptual object knowledge is housed, in part, in the same neural systems that subserve perception and action.

Ecphory of autobiographical memories: an fMRI study of recent and remote memory retrieval

Ecphory occurs when one recollects a past event cued by a trigger, such as a picture, odor, or name. It is a central component of autobiographical memory, which allows us to "travel mentally back in time" and re-experience specific events from our personal past. Using fMRI and focusing on the role of medial temporal lobe (MTL) structures, we investigated the brain bases of autobiographical memory and whether they change with the age of memories. Importantly, we used an ecphory task in which the remote character of the memories was ensured. The results showed that a large bilateral network supports autobiographical memory: temporal lobe, temporo-parieto-occipital junction, dorsal prefrontal cortex, medial frontal cortex, retrosplenial cortex and surrounding areas, and MTL structures. This network, including MTL structures, changed little with the age of the memories.

Activation of brain regions vulnerable to Alzheimer's disease: the effect of mild cognitive impairment

This study examined the functionality of the medial temporal lobe (MTL) and posterior cingulate (PC) in mild cognitive impairment amnestic type (MCI), a syndrome that puts patients at greater risk for developing Alzheimer disease (AD). Functional MRI (fMRI) was used to identify regions normally active during encoding of novel items and recognition of previously learned items in a reference group of 77 healthy young and middle-aged adults. The pattern of activation in this group guided further comparisons between 14 MCI subjects and 14 age-matched controls. The MCI patients exhibited less activity in the PC during recognition of previously learned items, and in the right hippocampus during encoding of novel items, despite comparable task performance to the controls. Reduced fMRI signal change in the MTL supports prior studies implicating the hippocampus for encoding new information. Reduced signal change in the PC converges with recent research on its role in recognition in normal adults as well as metabolic decline in people with genetic or cognitive risk for AD. Our results suggest that a change in function in the PC may account, in part, for memory recollection failure in AD.

Mining the posterior cingulate: Segregation between memory and pain components

We present a general method for automatic meta-analyses in neuroscience and apply it on text data from published functional imaging studies to extract main functions associated with a brain area-the posterior cingulate cortex (PCC). Abstracts from PubMed are downloaded, words extracted and converted to a bag-of-words matrix representation. The combined data are analyzed with hierarchical non-negative matrix factorization. We find that the prominent themes in the PCC corpus are episodic memory retrieval and pain. We further characterize the distribution in PCC of the Talairach coordinates available in some of the articles. This shows a tendency to functional segregation between memory and pain components where memory activations are predominantly in the caudal part and pain in the rostral part of PCC.

Task-dependent posterior cingulate activation in mild cognitive impairment

Neuroimaging research has demonstrated that the posterior cingulate cortex (PCC) is functionally compromised in individuals diagnosed with amnestic mild cognitive impairment (MCI), a major risk factor for the development of Alzheimer's disease (AD). In functional MRI studies with healthy participants, this same region is active during self-appraisal (requiring retrieval of semantic knowledge about the self) as well as episodic recognition of previously learned information. Administering both types of tasks to people with MCI may reveal important information on the role of the PCC in recollection. This study investigated fMRI activation in the PCC in individuals with MCI and matched controls across two tasks. The first task was a visual episodic recognition task. The second task was an autobiographical self-appraisal task in which subjects rated themselves on a set of trait adjectives. Results of a conjunction analysis revealed the PCC as the sole region commonly active during both tasks in the healthy older adults. Furthermore, additional analysis revealed an interaction in the PCC, indicating a task-dependent response in the MCI group. MCI participants showed PCC activation during self-appraisal, but not episodic retrieval. This result suggests in MCI that the PCC shows functional degradation during episodic retrieval; however, the PCC's role in retrieval and evaluation of highly elaborated information regarding the self is more well-preserved.

Overlap in the functional neural systems involved in semantic and episodic memory retrieval

Neuroimaging and neuropsychological data suggest that episodic and semantic memory may be mediated by distinct neural systems. However, an alternative perspective is that episodic and semantic memory represent different modes of processing within a single declarative memory system. To examine whether the multiple or the unitary system view better represents the data we conducted a network analysis using multivariate partial least squares (PLS ) activation analysis followed by covariance structural equation modeling (SEM) of positron emission tomography data obtained while healthy adults performed episodic and semantic verbal retrieval tasks. It is argued that if performance of episodic and semantic retrieval tasks are mediated by different memory systems, then there should differences in both regional activations and interregional correlations related to each type of retrieval task, respectively. The PLS results identified brain regions that were differentially active during episodic retrieval versus semantic retrieval. Regions that showed maximal differences in regional activity between episodic retrieval tasks were used to construct separate functional models for episodic and semantic retrieval. Omnibus tests of these functional models failed to find a significant difference across tasks for both functional models. The pattern of path coefficients for the episodic retrieval model were not different across tasks, nor were the path coefficients for the semantic retrieval model. The SEM results suggest that the same memory network/system was engaged across tasks, given the similarities in path coefficients. Therefore, activation differences between episodic and semantic retrieval may ref lect variation along a continuum of processing during task performance within the context of a single memory system.

Cortical representations of personally familiar objects and places: functional organization of the human posterior cingulate cortex

The recognition of both personally familiar objects and places involves nonspatial memory retrieval processes, but only personally familiar places are represented as space. Although the posterior cingulate cortex (PCC) is considered to process both types of such memories, its functional organization is poorly understood. In this event-related fMRI study, normal subjects judged familiar/unfamiliar pictures in four categories: familiar places (FP), familiar objects (FO), unfamiliar places (UP), and unfamiliar objects (UO), thus constituting a two-factorial design. A significant main effect of stimuli with greater activation in the place (FP and UP) than object (FO and UO) trials was observed bilaterally in several medial temporo-occipito-parietal regions, including the caudal PCC (cPCC) and parahippocampal gyrus. The reverse comparison revealed greater activation in the lateral inferior occipito-temporal junctions and intraparietal sulci bilaterally. A significant main effect of familiarity with greater activation in the familiar (FP and FO) than unfamiliar (UP and UO) trials was observed in the mid-dorsal PCC (mPCC), retrosplenial cortex, posterior precuneus, and the left intraparietal sulcus. Activation specific to the FP trials (as assessed by the interaction) was observed in the right posterodorsal PCC (pPCC) only. Together with data from previous functional imaging studies, the results suggest a functional segregation of human PCC with differential involvement of pPCC in spatial representations of personally familiar places and of the mPCC and retrosplenial cortex in episodic retrieval of personally familiar places and objects. Activation of the left intraparietal sulcus may reflect retrieval of memories related to object manipulation.

Age-dependent brain activation during forward and backward digit recall revealed by fMRI

In this study, brain activation associated with forward and backward digit recall was examined in healthy old and young adults using functional MRI. A number of areas were activated during the recall. In young adults, greater activation was found in the left prefrontal cortex (BA9) and the left occipital visual cortex during backward digit recall than forward digit recall. In contrast, the activation in the right inferior frontal gyrus (BA 44/45) was more extensive in forward digit recall than in backward digit recall. In older adults, backward recall generated stronger activation than forward recall in most areas, including the frontal, the parietal, the occipital, and the temporal cortices. In the backward recall condition, the right inferior frontal gyrus (BA44/45) showed more activation in the old group than in the young group. These results suggest that different neural mechanisms may be involved in forward and backward digit recall and brain functions associated with these two types of recall are differentially affected by aging.

Language in context: emergent features of word, sentence, and narrative comprehension

Context exerts a powerful effect on cognitive performance and is clearly important for language processing, where lexical, sentential, and narrative contexts should differentially engage neural systems that support lexical, compositional, and discourse level semantics. Equally important, but thus far unexplored, is the role of context within narrative, as cognitive demands evolve and brain activity changes dynamically as subjects process different narrative segments. In this study, we used fMRI to examine the impact of context, comparing responses to a single, linguistically matched set of texts when these were differentially presented as random word lists, unconnected sentences and coherent narratives. We found emergent, context-dependent patterns of brain activity in each condition. Perisylvian language areas were always active, consistent with their supporting core linguistic computations. Sentence processing was associated with expanded activation of the frontal operculum and temporal poles. The same stimuli presented as narrative evoked robust responses in extrasylvian areas within both hemispheres, including precuneus, medial prefrontal, and dorsal temporo-parieto-occipital cortices. The right hemisphere was increasingly active as contextual complexity increased, maximal at the narrative level. Furthermore, brain activity was dynamically modulated as subjects processed different narrative segments: left hemisphere activity was more prominent at the onset, and right hemisphere more prominent at the resolution of a story, at which point, it may support a coherent representation of the narrative as a whole. These results underscore the importance of studying language in an ecologically valid context, suggesting a neural model for the processing of discourse.

Temporal lobe activations of "feeling-of-knowing" induced by face-name associations

"Feeling-of-Knowing" (FOK) refers to the sense of what one knows and is a component of the human metamemory system. We investigated the neural correlates of the FOK induced by face-name associations using the Recall-Judgment-Recognition paradigm. Data were gathered using event-related functional magnetic resonance imaging (fMRI). We analyzed the fMRI data with parametric analyses of six FOK ratings while compensating for the effects of unbalanced response latencies among trials ("variable duration parametric analysis"). Regions showing a significant linear relationships with the FOK ratings (FOK regions) were the bilateral ventral, dorsal, and anterior prefrontal regions; the medial frontal regions; the medial surface regions; the left parietal regions; the bilateral superior temporal and nearby regions; the right anterior temporal region; and the bilateral thalami/basal ganglia. Most of the active areas in the prefrontal regions were common to those found in our previous FOK studies of general knowledge (Kikyo, H., Ohki, K., Miyashita, Y., 2002. Neural correlates for feeling-of-knowing: an fMRI parametric analysis. Neuron 36, 177-186). However, in this study, we found robust activations of the temporal regions near to the regions that were related to the higher-order information processing of face images or semantic information processing of the to-be-recalled person. Those results suggest that the information related to the higher-order visual features of a face, which was represented in the temporal cortex, was activated by the top-down signal from the prefrontal cortex, and that this cooperation between the temporal and prefrontal cortices may contribute to the FOK.

Puzzling thoughts for H. M.: can new semantic information be anchored to old semantic memories?

Researchers currently debate whether new semantic knowledge can be learned and retrieved despite extensive damage to medial temporal lobe (MTL) structures. The authors explored whether H. M., a patient with amnesia, could acquire new semantic information in the context of his lifelong hobby of solving crossword puzzles. First, H. M. was tested on a series of word-skills tests believed important in solving crosswords. He also completed 3 new crosswords: 1 puzzle testing pre-1953 knowledge, another testing post-1953 knowledge, and another combining the 2 by giving postoperative semantic clues for preoperative answers. From the results, the authors concluded that H. M. can acquire new semantic knowledge, at least temporarily, when he can anchor it to mental representations established preoperatively.

Medial temporal lobe structures are needed to re-experience remote autobiographical memories: evidence from H.M. and W.R

The nature and extent of retrograde amnesia in patients with medial temporal lobe (MTL) lesions is currently under debate. While some investigators propose a temporally limited role for the MTL in episodic and semantic memory, others claim that MTL structures are needed for episodic memories of one's entire lifetime, and that only semantic memory becomes independent of the MTL. To address this issue, we tested two amnesic patients, H.M. and W.R., with bilateral MTL lesions on a series of remote memory tests that together distinguished episodic memory from semantic memory performance. Notably, we used a new method to assess autobiographical memory that measured the degree of re-experiencing of personal happenings from the past. Both patients showed relatively spared semantic memory, but severe impairment on measures of autobiographical memory, with no temporal gradient. Our data support the view that MTL structures play a significant role in recalling specific personal episodes, not only from the recent past but from the distant past as well.

The mental self

In meditation both the quality and the contents of consciousness may be voluntarily changed, making it an obvious target in the quest for the neural correlate of consciousness. Here we present the results of a positron emission tomography study of yoga nidra relaxation meditation when compared with the normal resting conscious state. Meditation is accompanied by a relatively increased perfusion in the sensory imagery system: hippocampus and sensory and higher order association regions, with decreased perfusion in the executive system: dorsolateral prefrontal cortex, anterior cingulate gyrus, striatum, thalamus, pons, and cerebellum. To identify regions active in both systems we performed a principal component analysis of the results. This separated the blood flow data into two groups of regions, explaining 25 and 18% of their variance: One group corresponded to the executive system, and the other to the systems supporting sensory imagery. A small group of regions contributed considerably to both networks: medial parietal and medial prefrontal cortices, together with the striatum. The inclusion of the striatum and our subsequent finding of increased striatal dopamine binding to D2 receptors during meditation suggested dopaminergic regulation of this circuit. We then investigated the neural networks supporting episodic retrieval of judgments of individuals with different degrees of self-relevance, in the decreasing order: self, best friend, and the Danish queen. We found that all conditions activated a medial prefrontal - precuneus/posterior cingulate cortex, thalamus, and cerebellum. This activation occurred together with the activation of the left lateral prefrontal/temporal cortex. The latter was dependent on the requirement of retrieval of semantic information, being most pronounced in the "queen" condition. Transcranial magnetic stimulation, targeting precuneus, was then applied to the medial parietal region to transiently disrupt the normal function of the circuitry. We found a decreased efficiency of retrieval of self-judgment compared to the judgment of best friend. This shows that the integrity of the function of precuneus is essential for self-reference, but not for reference to others.

Neural systems for sign language production: mechanisms supporting lexical selection, phonological encoding, and articulation

Overt production of ASL signs was evaluated using H(2)(15)O PET to differentiate brain systems that support sign language production at the lexical-selection and phonological-articulatory levels. Subjects were 16 right-handed, congenitally deaf native ASL signers (10 women, six men; age 20 to 29 years). Scans were performed while subjects (1) passively viewed ASL nouns, (2) repeated nouns, (3) generated verbs in response to these nouns, (4) passively viewed videotaped segments depicting transitive actions, and (5) generated a verb to describe these actions. Conjunctions between the two verb-generation tasks revealed left-lateralized activation of perisylvian, frontal, and subcortical regions commonly observed in spoken language generation tasks and implicated in processes of semantic feature binding and lexical selection. Analysis of noun repetition minus viewing condition revealed activation of distinct systems supporting phonological encoding and articulation, including bilateral activation of sensorimotor areas and association cortices in the temporal, parietal, and occipital lobes. In addition, lexical-selection and articulatory processes were associated with activation of different corticostriatal-thalamocortical circuits: articulation with activation of the motor, and lexical-selection with activation of the prefrontal circuits, respectively. The results collectively provide insight into dissociable neural systems underlying these psycholinguistic functions. In addition, activation of regions that are typically associated with the auditory system during sign production suggests that these regions may support modality-independent linguistic processes, or may indicate cross-modal plasticity within the deaf brain.

Parietal cortex and representation of the mental Self

For a coherent and meaningful life, conscious self-representation is mandatory. Such explicit "autonoetic consciousness" is thought to emerge by retrieval of memory of personally experienced events ("episodic memory"). During episodic retrieval, functional imaging studies consistently show differential activity in medial prefrontal and medial parietal cortices. With positron-emission tomography, we here show that these medial regions are functionally connected and interact with lateral regions that are activated according to the degree of self-reference. During retrieval of previous judgments of Oneself, Best Friend, and the Danish Queen, activation increased in the left lateral temporal cortex and decreased in the right inferior parietal region with decreasing self-reference. Functionally, the former region was preferentially connected to medial prefrontal cortex, the latter to medial parietal. The medial parietal region may, then, be conceived of as a nodal structure in self-representation, functionally connected to both the right parietal and the medial prefrontal cortices. To determine whether medial parietal cortex in this network is essential for episodic memory retrieval with self-representation, we used transcranial magnetic stimulation over the region to transiently disturb neuronal circuitry. There was a decrease in the efficiency of retrieval of previous judgment of mental Self compared with retrieval of judgment of Other with transcranial magnetic stimulation at a latency of 160 ms, confirming the hypothesis. This network is strikingly similar to the network of the resting conscious state, suggesting that self-monitoring is a core function in resting consciousness.

Dissociating person-specific from general semantic knowledge: roles of the left and right temporal lobes

The cognitive architecture and neural underpinnings of different semantic domains remains highly controversial. We report two patients with focal temporal lobe atrophy who presented with contrasting and theoretically informative dissociations of person-specific versus general semantic knowledge. Subject J.P. showed severely impaired person-specific semantics, with relative preservation of knowledge about objects and animals, while subject M.A. exhibited the opposite pattern of performance (good knowledge of people in the context of impoverished general semantics). Voxel-based morphometric analysis of MR images in the two cases established predominantly right temporal atrophy associated with J.P.'s deficit for person knowledge and predominantly left temporal atrophy in M.A. who was impaired in general conceptual knowledge.

[Neural correlates of "negative" formal thought disorder]

Formal thought disorder (FTD) is a core feature of schizophrenia, but its pathophysiology is poorly understood. It can be conceptualised as 'positive' (e.g. incoherence, neologisms) or 'negative' (e.g. poverty of thought) thought disorder. We investigated the neural correlates of negative FTD. Blood oxygenation level-dependent (BOLD) contrast was measured using functional magnetic resonance imaging (fMRI) while six patients with schizophrenia and six healthy control subjects spoke about seven Rorschach inkblots for 3 min each. Varying degrees of thought-disordered speech were elicited during each run. In a within-subject design, the degree of negative thought disorder, conceptualised as 'poverty of speech' (Liddle et al. (2002) Thought and Language Index. Br J Psychiatry) was correlated with BOLD contrast in the two runs per patient showing the highest variance in this phenomenon. The degree of poverty of speech correlated positively mainly with activation in the right inferior parietal lobe (BA 40), middle frontal gyrus (BA 46), cuneus (BA 18), and the left posterior cingulate (BA 31). Negative correlations were evident in the left hippocampal/fusiform gyrus (BA 35/36/37). The severity of negative FTD correlates with activation in areas previously implicated in autobiographic episodic memory. During symptoms behaviourally characterised as negative, patients may experience rich memories and associations.

Macaque monkey retrosplenial cortex: II. Cortical afferents

We investigated the cortical afferents of the retrosplenial cortex and the adjacent posterior cingulate cortex (area 23) in the macaque monkey by using the retrograde tracers Fast blue and Diamidino yellow. We quantitatively analyzed the distribution of labeled neurons throughout the cortical mantle. Injections involving the retrosplenial cortex resulted in labeled neurons within the retrosplenial cortex and in areas 23 and 31 (approximately 78% of the total labeled cells). In the remainder of the cortex, the heaviest projections originated in the hippocampal formation, including the entorhinal cortex, subiculum, presubiculum, and parasubiculum. The parahippocampal and perirhinal cortices also contained many labeled neurons, as did the prefrontal cortex, mainly in areas 46, 9, 10, and 11, and the occipital cortex, mainly area V2. Injections in area 23 also resulted in numerous labeled cells in the posterior cingulate and retrosplenial regions (approximately 67% of total labeled cells). As in the retrosplenial cortex, injections of area 23 led to many labeled neurons in the frontal cortex, although most of these cells were in areas 9 and 46. Larger numbers of retrogradely labeled cells were also distributed more widely in the posterior parietal cortex, including areas 7a, 7m, LIP, and DP. There were some labeled cells in the parahippocampal cortex. These connections are consistent with the retrosplenial cortex acting as an interface between the working memory functions in the prefrontal areas and the long-term memory encoding in the medial temporal lobe. The posterior cingulate cortex, in contrast, may be more highly associated with visuospatial functions.

The cerebellum and olfaction in the aging brain: a functional magnetic resonance imaging study

The present study investigated activation of distinct cerebellar regions as a result of olfactory stimulation in healthy young and elderly adults using functional magnetic resonance imaging (fMRI). Ten young and 10 elderly adults were imaged using a 1.5-TMR scanner. The odorant amyl acetate was delivered in 12-s on, 40-s off cycles. Throughout the scan participants responded with a button press at first detection of each stimulus interval followed by a second button press upon odorant extinction. Images were processed with AFNI software. Elderly participants showed significantly decreased cerebellar activation in both the superior semilunar lobule (Crus I) and the inferior semilunar lobule (Crus II), two of three previously identified regions of interest for odor processing, compared to young adults. Interestingly, both groups showed similar levels of activation in the third region of interest, the posterior quadrangular lobule (VI), although the elderly showed more variability than the young did. Previous research identifying this area to be involved in attention may reflect the possibility that elderly adults engaged in olfactory tasks may show more variability than young participants in the degree of attentional demands needed for these tasks as a result of decreased olfactory abilities.

Multiple visual memory phenomena in a memory search task

This paper reports evidence of the existence of multiple and distinct visual memory processes in a memory search task in which a divided field stimulus presentation was used at study (Experiments 1-3) and either a foveal (Experiments 1 and 2) or a lateralized (Experiment 3) stimulus presentation was used at test. These memory processes can be distinguished on the basis of (1) whether or not they are hemispherically organized; and (2) the locus of their underlying brain activity, as evidenced by the scalp distribution of the event-related brain potentials and by the localization of the event-related optical signal that accompany them. These memory effects are discussed in the context of visual form memory.

Cortical Analysis of Visual Context

Objects in our environment tend to be grouped in typical contexts. How does the human brain analyze such associations between visual objects and their specific context? We addressed this question in four functional neuroimaging experiments and revealed the cortical mechanisms that are uniquely activated when people recognize highly contextual objects (e.g., a traffic light). Our findings indicate that a region in the parahippocampal cortex and a region in the retrosplenial cortex together comprise a system that mediates both spatial and nonspatial contextual processing. Interestingly, each of these regions has been identified in the past with two functions: the processing of spatial information and episodic memory. Attributing contextual analysis to these two areas, instead, provides a framework for bridging between previous reports.

Evidence from functional neuroimaging of a compensatory prefrontal network in Alzheimer's disease

Previous experiments have found that individuals with Alzheimer's disease (AD) show increased activity in prefrontal regions compared with healthy age-matched controls during cognitive tasks. This has been interpreted as compensatory reallocation of cognitive resources, but direct evidence for a facilitating effect on performance has been lacking. To address this we measured neural activity during semantic and episodic memory tasks in mildly demented AD patients and healthy elderly controls. Controls recruited a left hemisphere network of regions, including prefrontal and temporal cortices in both the semantic and episodic tasks. Patients engaged a unique network involving bilateral dorsolateral prefrontal and posterior cortices. Critically, activity in this network of regions was correlated with better performance on both the semantic and episodic tasks in the patients. This provides the most direct evidence to date that AD patients can use additional neural resources in prefrontal cortex, presumably those mediating executive functions, to compensate for losses attributable to the degenerative process of the disease.

Brain correlates of performance in a free/cued recall task with semantic encoding in Alzheimer disease

The goal of this study was to explore in patients with Alzheimer's disease (AD) the brain correlates of free and cued recall performance using an adaptation of the procedure developed by Grober and Buschke (1987). This procedure, which ensures semantic processing and coordinates encoding and retrieval, has been shown to be very sensitive to an early diagnosis of AD. Statistical parametric mapping (SPM 99) was used to establish clinicometabolic correlations between performance at free and cued verbal recall and resting brain metabolism in 31 patients with AD. Results showed that patient's score on free recall correlated with metabolic activity in right frontal regions (BA 10 and BA 45), suggesting that performance reflected a strategic retrieval attempt. Poor retrieval performance was tentatively attributed to a loss of functional correlation between frontal and medial temporal regions in patients with AD compared with elderly controls. Performance on cued recall was correlated to residual metabolic activity in bilateral parahippocampal regions (BA 36), suggesting that performance reflected retrieval of semantic associations, without recollection in AD. In conclusion, this study demonstrates that the diagnostic sensitivity for Alzheimer's disease of the cued recall performance in the Grober and Buschke procedure (1987) depends on the activity of parahippocampal regions, one of the earliest targets of the disease. Moreover, the results suggest that the poor performance of patients with AD during free and cued recall is related to a decreased connectivity between parahippocampal regions and frontal areas.

Functional imaging of the semantic system: retrieval of sensory-experienced and verbally learned knowledge

This paper considers how functional neuro-imaging can be used to investigate the organization of the semantic system and the limitations associated with this technique. The majority of the functional imaging studies of the semantic system have looked for divisions by varying stimulus category. These studies have led to divergent results and no clear anatomical hypotheses have emerged to account for the dissociations seen in behavioral studies. Only a few functional imaging studies have used task as a variable to differentiate the neural correlates of semantic features more directly. We extend these findings by presenting a new study that contrasts tasks that differentially weight sensory (color and taste) and verbally learned (origin) semantic features. Irrespective of the type of semantic feature retrieved, a common semantic system was activated as demonstrated in many previous studies. In addition, the retrieval of verbally learned, but not sensory-experienced, features enhanced activation in medial and lateral posterior parietal areas. We attribute these "verbally learned" effects to differences in retrieval strategy and conclude that evidence for segregation of semantic features at an anatomical level remains weak. We believe that functional imaging has the potential to increase our understanding of the neuronal infrastructure that sustains semantic processing but progress may require multiple experiments until a consistent explanatory framework emerges.

Retrieval of relational information: a role for the left inferior prefrontal cortex

Neuroimaging studies have implicated different areas of prefrontal cortex and medial temporal lobe structures (MTL) in episodic retrieval tasks. However, the role of specific regions in particular aspects of episodic memory is still unclear. In this experiment we studied changes in regional cerebral blood flow (rCBF) associated with relational and nonrelational retrieval of studied pairs of words. For relational retrieval, a list of either studied or rearranged pairs was presented and subjects (n = 8) were asked to indicate whether pairs had appeared on the study list. Under the nonrelational retrieval condition they indicated whether one or both words of the pair had appeared on the study list. As compared to the baseline condition (looking at a cross-mark), increased rCBF was observed in the left inferior prefrontal cortex (LIPFC) for both studied pairs and rearranged pairs under the relational retrieval condition. Under the nonrelational condition, an increase was observed in right inferior frontal gyrus. The MTL showed a trend for increased rCBF in the rearranged-pair condition. This increase was probably associated with the encoding that accompanies retrieval of novel stimuli. Results suggest that the lateralized activation of prefrontal cortex observed in episodic memory tasks may be related to the degree of relational processing involved. The LIPFC appears to be associated with relational retrieval and the right prefrontal cortex with nonrelational retrieval.

Is knowledge of famous people disproportionately impaired in patients with early and questionable Alzheimer's disease?

Twenty-two patients with early dementia of Alzheimer's type (DAT) and 31 controls were administered tests of person-specific semantics (Experiment 1). DAT patients were impaired on all test components. In Experiment 2, 31 DAT patients, 28 questionable DAT (QDAT) patients, and 42 controls were administered the Graded Naming Test (GNT) and the newly designed Graded Faces Test (GFT), matched for difficulty with the GNT. DAT patients were impaired throughout but showed an advantage for naming objects over faces. The QDAT patients were impaired on the GFT only. Of the 7 QDAT patients who evolved to DAT within 1-2 years, 6 showed initial impairment on the GFT, whereas 17 of the nonconverters scored normally on the GFT. Results suggest greater and earlier vulnerability of person knowledge than general semantic knowledge in DAT.

"Pray or Prey?" dissociation of semantic memory retrieval from episodic memory processes using positron emission tomography and a novel homophone task

One problem in studying the neural basis of semantic memory using functional neuroimaging is that it is often difficult to disentangle activation associated with semantic memory retrieval from that associated with episodic memory encoding and retrieval. To address this issue, a novel homophone task was used in which subjects were PET scanned whilst learning a series of real words (e.g., prey). In a subsequent scan, the subjects were presented with homophone pairs (e.g., prey vs pray) and were required to choose the one that had been shown previously. In two corresponding baseline tasks, the subjects were scanned whilst learning and recognizing pronounceable nonwords. Thus, while all of these tasks recruited either episodic memory encoding or retrieval processes, only the homophone tasks involved semantic memory retrieval. A conjunction analysis designed to isolate activation associated with semantic memory retrieval, revealed changes in several left lateral frontal regions (BA 9/10, 9/45), the left middle temporal cortex (BA 21), and in the left inferior temporoparietal cortex (BA 39). In contrast, a conjunction analysis designed to isolate activation associated with episodic memory encoding, revealed significant changes in the left hippocampus, as well as in the frontopolar cortex (BA 10) bilaterally, the left inferior parietal cortex (BA 40), and the left superior temporal gyrus (BA 22, 28). The present results clarify and extend recent attempts to understand the neural basis of semantic memory retrieval, by actively controlling for the confounding effects of episodic memory encoding and retrieval processes.

The neural substrates of episodic memory impairment in Alzheimer's disease as revealed by FDG-PET: relationship to degree of deterioration

In a previous investigation, we raised the hypothesis that in Alzheimer's disease the cerebral structures implicated in episodic memory deficits may differ according to the severity of cognitive impairment. To test this hypothesis, Story Recall test scores and PET measurements of resting cerebral glucose utilization, a measure of synaptic integrity, were obtained in 40 patients. Using SPM96 (statistical parametric mapping 1996), positive correlations between the two sets of data were calculated on a voxel basis, first in the whole patient sample and then separately in the two subgroups of 20 patients differing in Mini-Mental State Examination score, i.e. those with least impaired and those with most impaired performance ('less severe' and 'more severe' subgroups, respectively). In the whole sample, significant correlations (P < 0.05, corrected for multiple tests) involved bilaterally not only several limbic structures (the hippocampal/rhinal cortex regions, posterior cingulate gyrus and retrosplenial cortex) but also, and less expectedly, some temporo-occipital association areas. However, the subgroup analysis disclosed that, in the less severe subgroup, all significant correlations (P < 0.005, uncorrected) were restricted to the parahippocampal gyrus and retrosplenial cortex, in accordance with both the distribution of changes in tau in early Alzheimer's disease and the known involvement of this network in normal and impaired memory function, while in the more severe subgroup they mainly involved the left temporal neocortex, which is known to be implicated in semantic memory. These findings suggest that, when episodic memory is mildly impaired, limbic functions are still sufficient to subserve the remaining performance, whereas with more severe memory deficit resulting from accumulated pathology the neocortical areas that are normally involved in semantic memory are recruited, perhaps as a form of (inadequate) compensatory mechanism.

Brain imaging of human memory systems: between-systems similarities and within-system differences

There is much evidence for the existence of multiple memory systems. However, it has been argued that tasks assumed to reflect different memory systems share basic processing components and are mediated by overlapping neural systems. Here we used multivariate analysis of PET-data to analyze similarities and differences in brain activity for multiple tests of working memory, semantic memory, and episodic memory. The results from two experiments revealed between-systems differences, but also between-systems similarities and within-system differences. Specifically, support was obtained for a task-general working-memory network that may underlie active maintenance. Premotor and parietal regions were salient components of this network. A common network was also identified for two episodic tasks, cued recall and recognition, but not for a test of autobiographical memory. This network involved regions in right inferior and polar frontal cortex, and lateral and medial parietal cortex. Several of these regions were also engaged during the working-memory tasks, indicating shared processing for episodic and working memory. Fact retrieval and synonym generation were associated with increased activity in left inferior frontal and middle temporal regions and right cerebellum. This network was also associated with the autobiographical task, but not with living/non-living classification, and may reflect elaborate retrieval of semantic information. Implications of the present results for the classification of memory tasks with respect to systems and/or processes are discussed.

Retrieval of visual, auditory, and abstract semantics

Conceptual knowledge is thought to be represented in a large distributed network, indexing a range of different semantic features (e.g., visual, auditory, functional). We investigated the anatomical organization of these features, using PET, by contrasting brain activity elicited by heard words with (i) visual (e.g., blue), (ii) auditory (e.g., noise), or (iii) abstract (e.g., truth) meaning. The activation task was either repetition or semantic decision (e.g., does the meaning of the word relate to religion?). In the baseline conditions, the sound track of the words was reversed and subjects had to say "OK" (control for repetition) or make an acoustic decision (control for semantic decision). Irrespective of task, words relative to their corresponding controls activated the left posterior inferior temporal and inferior frontal cortices. In addition, semantic decisions on words with sensory (visual and auditory) meanings enhanced activation in a ventral region of the left anterior temporal pole. These results are consistent with neuropsychological studies showing that anterior temporal lobe damage can cause deficits for items that are mainly defined by their sensory features (i.e., concrete, particularly living items). Since modality-specific activation was observed only during the semantic decision task, we discuss whether it reflects retrieval of sensory semantics per se or the degree to which semantic associations are triggered during effortful retrieval.

The neural correlates of intentional and incidental self processing

The neuroscientific study of the 'Self' is just beginning to emerge. We used functional Magnetic Resonance Imaging (fMRI) to investigate cerebral activation while subjects processed words describing personality traits and physical features, in two experiments with contrasting designs: incidental and intentional. In the first experiment (intentional self processing), subjects were presented with personality trait adjectives and made judgements as to their self descriptiveness (versus non self descriptiveness). In the second experiment (incidental self processing), subjects categorised words according to whether they described physical versus psychological attributes, while unaware that the words had been arranged in blocks according to self descriptiveness. The subjects had previously rated all words for self descriptiveness 6 weeks prior to the scanning session. A reaction time advantage was present in both experiments for self descriptive trait words, suggesting a facilitation effect. Common areas of activation for the two experiments included the left superior parietal lobe, with adjacent regions of the lateral prefrontal cortex also active in both experiments. Differential signal changes were present in the left precuneus for the intentional and the right middle temporal gyrus for the incidental experiment. The results suggest that self processing involves distinct processes and can occur on more than one cognitive level with corresponding functional neuroanatomic correlates in areas previously implicated in the awareness of one's own state.

Stress, memory, and the hippocampus: can't live with it, can't live without it

Since the 1968s discovery of receptors for stress hormones (corticosteroids) in the rodent hippocampus, a tremendous amount of data has been gathered on the specific and somewhat isolated role of the hippocampus in stress reactivity. The hippocampal sensitivity to stress has also been extended in order to explain the negative impact of stress and related stress hormones on animal and human cognitive function. As a consequence, a majority of studies now uses the stress-hippocampus link as a working hypothesis in setting up experimental protocols. However, in the last decade, new data were gathered showing that stress impacts on many cortical and subcortical brain structures other than the hippocampus. The goal of this paper is to summarize the four major arguments previously used in order to confirm the stress-hippocampus link, and to describe new data showing the implication of other brain regions for each of these previously used arguments. The conclusion of this analysis will be that scientists should gain from extending the impact of stress hormones to other brain regions, since hormonal functions on the brain are best explained by their modulatory role on various brain structures, rather than by their unique impact on one particular brain region.

Large, colorful, or noisy? Attribute- and modality-specific activations during retrieval of perceptual attribute knowledge

Position emission tomography was used to investigate whether retrieval of perceptual knowledge from long-term memory activates unique cortical regions associated with the modality and/or attribute type retrieved. Knowledge about the typical color, size, and sound of common objects and animals was probed, in response to written words naming the objects. Relative to a nonsemantic control task, all the attribute judgments activated similar left temporal and frontal regions. Visual (color, size) knowledge selectively activated the right posterior inferior temporal (PIT) cortex, whereas sound judgments elicited selective activation in the left posterior superior temporal gyrus and the adjacent parietal cortex. All of the attribute judgments activated a left PIT region, but color retrieval generated more activation in this area. Size judgments activated the right medial parietal cortex. These results indicate that the retrieval of perceptual semantic information activates not only a general semantic network, but also cortical areas specialized for the modality and attribute type of the knowledge retrieved.

Transperceptual encoding and retrieval processes in memory: a PET study of visual and haptic objects

An important objective of functional neuroimaging research is to identify neuroanatomical correlates of memory processes such as encoding and retrieval. In typical studies directed at this goal, however, the to-be-remembered information has been presented in a single perceptual modality. Under these conditions it is not known whether the observed brain activity reflects the studied memory process as such or only the memory process in the given modality. The positron emission tomography (PET) study reported here was designed to identify brain regions involved in encoding and retrieval processes specific to visual and haptic modalities, as well as those common to the two modalities. These latter, common regions, were assumed to be associated with "transperceptual" encoding and retrieval processes. Abstract three-dimensional objects, difficult to describe verbally, served as to-be-remembered materials. A multivariate partial least squares analysis of the PET data revealed that transperceptual encoding processes activated right medial temporal lobe, superior prefrontal cortex bilaterally, and posterior inferior temporal gyrus bilaterally. Transperceptual recognition activations were observed in two right orbitofrontal regions and in anterior cingulate. These results provide initial evidence that some processes involved in memory encoding and retrieval operate beyond perceptual processes and in that sense are transperceptual.

A temporoparietal and prefrontal network for retrieving the spatial context of lifelike events

Virtual reality (VR) and event-related functional magnetic resonance imaging were used to study memory for the spatial context of controlled but lifelike events. Subjects received a set of objects from two different people in two different places within a VR environment. Memory for the objects, and for where and from whom they were received was tested by putting the subject back into a place in the company of a person and giving a paired forced choice of objects. In four conditions objects had to be chosen according to different criteria: which was received in that place, which was received from that person, which object was recognized, and which object was widest. Event-related functional magnetic resonance imaging was performed during testing to identify areas involved in retrieval of the spatial context of an event. A network of areas was identified consisting of a temporoparietal pathway running between the precuneus and parahippocampi via retrosplenial cortex and the parieto-occipital sulcus, left hippocampus, bilateral posterior parietal, dorsolateral, ventrolateral and anterior prefrontal cortices, and the anterior cingulate. Of these areas the parahippocampal, right posterior parietal, and posteriodorsal medial parietal areas were specifically involved in retrieval of spatial context compared to retrieval of nonspatial context. The posterior activations are consistent with a model of long-term storage of allocentric representations in medial temporal regions with translation to body-centered and head-centered representations computed in right posterior parietal cortex and buffered in the temporoparietal pathway so as to provide an imageable representation in the precuneus. Prefrontal activations are consistent with strategic retrieval processes, including those required to overcome the interference between the highly similar events.

Brain regions involved in prospective memory as determined by positron emission tomography

Prospective memory (PM) refers to the functions that enables a person to carry out an intended act after a delay. Despite the ubiquity of this behaviour, little is known about the supporting brain structures and the roles that they play. In this study, eight healthy participants performed four different PM tasks, each under three conditions: a baseline, and two conditions involving an intention. In the first of the intention conditions, subjects were asked to make a novel response to a certain class of stimuli whilst performing an attention-demanding task. However, the expected stimuli never actually occurred. In the second intention condition subjects were expecting to see these stimuli as before, and they did occur on approximately 20% of trials. Relative to the baseline condition, increases in regional cerebral blood flow (rCBF) as estimated by oxygen-15 positron emission tomography technique across all four tasks were seen in the frontal pole (Brodmann's area 10) bilaterally, right lateral prefrontal and inferior parietal regions plus the precuneus when subjects were expecting a PM stimulus regardless of whether it actually occurred. Further activation was seen in the thalamus when the PM stimuli occurred and was acted upon, with a corresponding rCBF decrease in right lateral prefrontal cortex. It is argued that the first set of region play a role in the maintenance of an intention, with the second set involved additionally in its realisation.

Conscious recollection and illusory recognition: an event-related fMRI study

In this event-related functional magnetic resonance imaging (fMRI) study we examined the neuronal correlates of the subprocesses underlying recognition memory. In an explicit memory task, participants had to discriminate studied ('old') words from semantically related and unrelated 'new' (unstudied) words. We examined whether the correct rejection of semantically related words which were similar to old words, which had elicited correct responses, was based on conscious recollection of study phase information. In this task, false-positive responses to semantically related new words can be assumed to result from the assessment of the semantic similarity between test words and studied words with minimal recollection. For correct identification of old words and correct rejection of new related words we found common activation in a variety of brain areas that have been shown to be involved in conscious recollection, among them the left middle frontal gyrus, the precuneus, the retrosplenial cortex, the left parahippocampal gyrus and the thalamus. For correct responses to old words, the frontomedian wall, the posterior cingulate cortex and the nucleus accumbens were additionally activated, suggesting an emotional contribution to these judgements. Correct rejections of related new words were associated with additional activation of the right middle frontal gyrus, suggesting higher monitoring demands for these more difficult recognition judgements. False-positive responses to semantically related new words were associated with enhanced activation in the frontomedian wall. The results point to an important role of the prefrontal cortex as well as medial temporal and medial parietal regions of the brain in successful memory retrieval and conscious recollection.

The role of color in high-level vision

Traditional theories of object recognition have emphasized the role of shape information in high-level vision. However, the accumulating behavioral, neuroimaging and neuropsychological evidence indicates that the surface color of an object affects its recognition. In this article, we discuss the research that examines the conditions under which color influences the operations of high-level vision and the neural substrates that might mediate these operations. The relationship between object color and object recognition is summarized in the 'Shape+Surface' model of high-level vision.

Semantic memory and the brain: structure and processes

Recent functional brain imaging studies suggest that object concepts may be represented, in part, by distributed networks of discrete cortical regions that parallel the organization of sensory and motor systems. In addition, different regions of the left lateral prefrontal cortex, and perhaps anterior temporal cortex, may have distinct roles in retrieving, maintaining and selecting semantic information.

A slow wave investigation of working memory biases in mood disorders

Mood-congruent working memory biases were examined in a delayed matching to sample paradigm using the slow wave (SW) event-related brain potential (ERP) component. Mood-congruent working memory biases, indexed by SW amplitudes, were demonstrated among individuals experiencing a major depressive episode (MDE) and nondepressed controls but not individuals with dysthymia. However, analyses of symptom severity demonstrated that those with dysthymia exhibited significantly less negative SW amplitudes with increasing depressive mood severity, whereas individuals with major depression demonstrated more negative SW amplitudes with increasing depressive mood severity. These results are discussed in the context of diagnostic specificity for cognitive biases associated with working memory of mood-disordered individuals.

Repetition related ERP effects in a visual object target detection task

ERP responses to initial and repeated presentations of possible and impossible objects were recorded from 61 recording sites in a simple target detection task. In Experiment 1, the non-target objects were line drawings of possible and impossible 3-D geometric figures and the targets were line drawings of familiar everyday objects or combinations of parts of everyday objects. In Experiment 2, the non-target objects were everyday objects and the targets were possible and impossible 3-D geometric figures. In both experiments, at frontal sites, the repeated possible and impossible non-target items elicited less negative ERP waveforms relative to first presentations between 250 and 350-400 ms. At parieto-occipital sites, in both experiments, the repeated possible and impossible non-target items elicited less positive ERP waveforms than did first presentations beginning at about 300 ms. The briefly reduced frontal negativity to repeated items is consistent with familiarity arising from a facilitation of access to conceptual, semantic and visuo-spatial representations during object categorization. The polarity of the parieto-occipital effect was the reverse of what is usually found in stimulus repetition tasks, although it is consistent with earlier work using similar visual stimuli. It is interpreted as reflecting the availability of a newly formed representation (i.e., token) of the object just experienced.

Episodic encoding and recognition of pictures and words: role of the human medial temporal lobes

In the present PET study, we examined brain activity related to processing of pictures and printed words in episodic memory. Our goal was to determine how the perceptual format of objects (verbal versus pictorial) is reflected in the neural organization of episodic memory for common objects. We investigated this issue in relation to encoding and recognition with a particular focus on medial temporal-lobe (MTL) structures. At encoding, participants saw pictures of objects or their written names and were asked to make semantic judgments. At recognition, participants made yes-no recognition judgments in four different conditions. In two conditions, target items were pictures of objects; these objects had originally been encoded either in picture or in word format. In two other conditions, target items were words; they also denoted objects originally encoded either as pictures or as words. Our data show that right MTL structures are differentially involved in picture processing during encoding and recognition. A posterior MTL region showed higher activation in response to the presentation of pictures than of words across all conditions. During encoding, this region may be involved in setting up a representation of the perceptual information that comprises the picture. At recognition, it may play a role in guiding retrieval processes based on the perceptual input, i.e. the retrieval cue. Another more anterior right MTL region was found to be differentially involved in recognition of objects that had been encoded as pictures, irrespective of whether the retrieval cue provided was pictorial or verbal in nature; this region may be involved in accessing stored pictorial representations. Our results suggest that left MTL structures contribute to picture processing only during encoding. Some regions in the left MTL showed an involvement in semantic encoding that was picture specific; others showed a task-specific involvement across pictures and words. Together, our results provide evidence that the involvement of some but not all MTL regions in episodic encoding and recognition is format specific.

Combined semantic dementia and apraxia in a patient with frontotemporal lobar degeneration

In this study we report neuropsychological and brain-imaging findings in a patient with frontotemporal lobar degeneration. Brain imaging using registration of (18)fluorodeoxyglucose (FDG)-PET data to three-dimensional (3-D) magnetic resonance imaging showed atrophy and highly significant hypometabolism of the left temporal lobe and both frontal lobes. Volumetric measurements of the hippocampi/amygdala showed a reduction in volume of 25% on the left compared to right within cortical areas. Neuropsychological testing revealed semantic dementia with severe anomia as well as apraxia with impairment of both recognition and production of motor acts. The implications of this case of early manifestation of frontotemporal lobar degeneration for our knowledge of dementia are discussed.