Redefining the functional organization of working memory processes within human lateral prefrontal cortex

It is widely held that the frontal cortex plays a critical part in certain aspects of spatial and non-spatial working memory. One unresolved issue is whether there are functionally distinct subdivisions of the lateral frontal cortex that subserve different aspects of working memory. The present study used positron emission tomography (PET) to demonstrate that working memory processes within the human mid-dorsolateral and mid-ventrolateral frontal regions are organized according to the type of processing required rather than according to the nature (i.e. spatial or non-spatial), of the information being processed, as has been widely assumed. Two spatial working memory tasks were used which varied in the extent to which they required different executive processes. During a 'spatial span' task that required the subject to hold a sequence of five previously remembered locations in working memory a significant change in blood-flow was observed in the right mid-ventrolateral frontal cortex, but not in the anatomically and cytoarchitectonically distinct mid-dorsolateral frontal-lobe region. By contrast, during a '2-back' task that required the subject to continually update and manipulate an ongoing sequence of locations within working memory, significant blood flow increases were observed in both mid-ventrolateral and mid-dorsolateral frontal regions. When the two working memory tasks were compared directly, the one that emphasized manipulation of information within working memory yielded significantly greater activity in the right mid-dorsolateral frontal cortex only. This dissociation provides unambiguous evidence that the mid-dorsolateral and mid-ventrolateral frontal cortical areas make distinct functional contributions to spatial working memory and corresponds with a fractionation of working memory processes in psychological terms.     

Disordered integration of heteromodal short-term cognitive operations: A breakdown of working memory

Abstract

We report a patient with bilateral frontal lobe infarcts with a symptom that has not yet been documented in the literature. He complained of severe difficulty in making a phone call. On examination, he demonstrated extreme difficulty in pointing to a series of numbers, although he had normal vision, normal motor and sensory functions, fair auditory memory for numbers and memory for spatial span. Based on a number of tests, we concluded that the patient's principal disorder was difficulty in converting one mode of information into another mode of information. This unexpected symptom can be best explained in terms of a working memory paradigm. Although individual modal systems, or slave systems, retained nearly normal function, simultaneous holding of heteromodal information and its modal transformation in the central executive system seemed to have been at fault.     

Attention and working memory: a dynamical model of neuronal activity in the prefrontal cortex

Cognitive behaviour requires complex context-dependent mapping between sensory stimuli and actions. The same stimulus can lead to different behaviours depending on the situation, or the same behaviour may be elicited by different cueing stimuli. Neurons in the primate prefrontal cortex show task-specific firing activity during working memory delay periods. These neurons provide a neural substrate for mapping stimulus and response in a flexible, context- or rule-dependent, fashion. We describe here an integrate-and-fire network model to explain and investigate the different types of working-memory-related neuronal activity observed. The model contains different populations (or pools) of neurons (as found neurophysiologically) in attractor networks which respond in the delay period to the stimulus object, the stimulus position ('sensory pools'), to combinations of the stimulus sensory properties (e.g. the object identity or object location) and the response ('intermediate pools'), and to the response required (left or right) ('premotor pools'). The pools are arranged hierarchically, are linked by associative synaptic connections, and have global inhibition through inhibitory interneurons to implement competition. It is shown that a biasing attentional input to define the current rule applied to the intermediate pools enables the system to select the correct response in what is a biased competition model of attention. The integrate-and-fire model not only produces realistic spiking dynamicals very similar to the neuronal data but also shows how dopamine could weaken and shorten the persistent neuronal activity in the delay period; and allows us to predict more response errors when dopamine is elevated because there is less different activity in the different pools of competing neurons, resulting in more conflict.     

Frontal theta and beta synchronizations for monetary reward increase visual working memory capacity

Visual working memory (VWM) capacity is affected by motivational influences; however, little is known about how reward-related brain activities facilitate the VWM systems. To investigate the dynamic relationship between VWM- and reward-related brain activities, we conducted time–frequency analyses using electroencephalograph (EEG) data obtained during a monetary-incentive delayed-response task that required participants to memorize the position of colored disks. In case of a correct answer, participants received a monetary reward (0, 10 or 50 Japanese yen) announced at the beginning of each trial. Behavioral results showed that VWM capacity under high-reward condition significantly increased compared with that under low- or no-reward condition. EEG results showed that frontal theta (6 Hz) amplitudes enhanced during delay periods and positively correlated with VWM capacity, indicating involvement of theta local synchronizations in VWM. Moreover, frontal beta activities (24 Hz) were identified as reward-related activities, because delay-period amplitudes correlated with increases in VWM capacity between high-reward and no-reward conditions. Interestingly, cross-frequency couplings between frontal theta and beta phases were observed only under high-reward conditions. These findings suggest that the functional dynamic linking between VWM-related theta and reward-related beta activities on the frontal regions plays an integral role in facilitating increases in VWM capacity.     

Frontal theta activity in humans increases with memory load in a working memory task

Recent theoretical work has suggested that brain oscillations in the theta band are involved in active maintenance and recall of working memory representations. To test this theoretical framework we recorded neuromagnetic responses from 10 subjects performing the Sternberg task. Subjects were required to retain a list of 1, 3, 5 or 7 visually presented digits during a 3-s retention period. During the retention period we observed ongoing frontal theta activity in the 7-8.5-Hz band recorded by sensors over frontal brain areas. The activity in the theta band increased parametrically with the number of items retained in working memory. A time-frequency analysis revealed that the task-dependent theta was present during the retention period and during memory scanning. Following the memory task the theta activity was reduced. These results suggest that theta oscillations generated in frontal brain regions play an active role in memory maintenance.     

Excitotoxic lesions of the rhinal cortex in the baboon differentially affect visual recognition memory, habit memory and spatial executive functions

To specify the functional role of the rhinal cortex, baboons with bilateral excitotoxic lesions of the rhinal cortex (RH group) were tested on a series of computerized memory and learning tasks. Preoperatively, they were trained to and then tested on a delayed nonmatching-to-sample (DNMS) task with trial-unique stimuli. Postoperatively, this visual recognition memory task was given twice. As compared to a sham-operated group, the RH group showed good retention of rule learning and were unimpaired on the Delay memory subtest. Performance on the List Length memory subtest was, however, severely impaired at both postoperative evaluations, with a significant negative correlation between cognitive performance and neuronal loss in rhinal areas. Visual habit memory and spatial working memory were assessed postoperatively only, using a concurrent discrimination learning task and both a delayed-response task (with a two- and four-location choice) and a delayed alternation task, respectively. The RH group was unimpaired on the first two tasks and was even faster than the controls in learning the delayed-response task with four locations. Finally, most RH baboons failed to learn the delayed alternation task within the limits of testing. These results indicate that neuronal loss in the rhinal cortex is sufficient to impair visual recognition memory, and extend the implication of this area to spatial executive functions. Furthermore, the observation of impaired recognition memory and executive processes with preserved procedural memory and retrograde memory suggests that damage to the rhinal cortex probably participates in the cognitive deficits typical of the early stages of Alzheimer's disease.     

Attentional dysfunction of the central executive in AD: evidence from dual task and perseveration errors

This study examined the capacity of 27 Alzheimer's disease (AD) patients to divide attention between two simultaneous tasks, as compared to 27 elderly controls. In order to determine whether distribution of attention is affected by age, a younger group was included. The results showed a marked impairment in the capacity of the AD patients to combine performance in two simultaneous tasks compared to the elderly controls, but the latter group did not differ from the younger participants, indicating a disease rather than an age effect, and replicating the results of Baddeley et al. (2001). The present study also analysed whether perseverations committed in a previous study (Sebastian et al., 2001) by some AD and elderly control participants in the Brown-Peterson (B-P) task (Brown, 1958; Peterson and Peterson, 1959) were related to their attentional capacity in the dual task. Although the elderly controls committed perseveration errors, a negative correlation was only found in the AD group between the index of attention distribution in the dual task and the frequency of perseverations in the B-P task. Our results would therefore suggest that AD patients show a dysfunction of the central executive and/or a deficit in their ability to allocate cognitive resources. This could be interpreted in terms of problems in updating the contents of working memory and suppressing activation of no-longer-relevant information in the early stages of AD, leading to perseveration errors in the B-P task.     

Refreshing: a minimal executive function

Executive functions include processes by which important information (e.g., words, objects, task goals, contextual information) generated via perception or thought can be foregrounded and thereby influence current and subsequent processing. One simple executive process that has the effect of foregrounding information is refreshing--thinking briefly of a just-activated representation. Previous studies (e.g., Johnson et al., 2005) identified refresh-related activity in several areas of left prefrontal cortex (PFC). To further specify the respective functions of these PFC areas in refreshing, in Experiment 1, healthy young adult participants were randomly cued to think of a just previously seen word (refresh) or cued to press a button (act). Compared to simply reading a word, refresh and act conditions resulted in similar levels of activity in left lateral anterior PFC but only refreshing resulted in greater activity in left dorsolateral PFC. In Experiment 2, refreshing was contrasted with a minimal phonological rehearsal condition. Refreshing was associated with activity in left dorsolateral PFC and rehearsing with activity in left ventrolateral PFC. In both experiments, correlations of activity among brain areas suggest different functional connectivity for these processes. Together, these findings provide evidence that (1) left lateral anterior PFC is associated with initiating a non-automatic process, (2) left dorsolateral PFC is associated with foregrounding a specific mental representation, and (3) refreshing and rehearsing are neurally distinguishable processes.     

Dorsolateral prefrontal cortex, working memory and episodic memory processes: insight through transcranial magnetic stimulation techniques

The ability to recall and recognize facts we experienced in the past is based on a complex mechanism in which several cerebral regions are implicated. Neuroimaging and lesion studies agree in identifying the frontal lobe as a crucial structure for memory processes, and in particular for working memory and episodic memory and their relationships. Furthermore, with the introduction of transcranial magnetic stimulation (TMS) a new way was proposed to investigate the relationships between brain correlates, memory functions and behavior. The aim of this review is to present the main findings that have emerged from experiments which used the TMS technique for memory analysis. They mainly focused on the role of the dorsolateral prefrontal cortex in memory process. Furthermore, we present state-of-the-art evidence supporting a possible use of TMS in the clinic. Specifically we focus on the treatment of memory deficits in depression and anxiety disorders.     

Prefrontal cortex dysfunction mediates deficits in working memory and prepotent responding in schizophrenia.

BACKGROUND: Schizophrenic patients show deficits in working memory (WM) and inhibition of prepotent responses. We examined brain activity while subjects performed tasks that placed demands on WM and overriding prepotent response tendencies, testing predictions that both processes engage overlapping prefrontal cortical (PFC) regions and that schizophrenic patients show reduced PFC activity and performance deficits reflecting both processes. METHODS: Functional magnetic resonance imaging data were acquired while 16 schizophrenic and 15 healthy subjects performed the N-Back task that varied WM load and a version of the AX-CPT that required overriding a prepotent response tendency. RESULTS: Both tasks engaged overlapping cortical networks (e.g., bilateral dorsolateral PFC, Broca's area, parietal cortex). Increased WM load monotonically increased activity; preparation to override a prepotent response produced greater and more enduring activity. Group differences on each task emerged in a right dorsolateral PFC region: schizophrenic subjects showed lesser magnitude increases under conditions of high WM and prepotent response override demands, with concomitant performance impairments. CONCLUSIONS: Schizophrenic patients exhibit PFC-mediated deficits in WM and preparation to override prepotent responses. Findings are consistent with the operation of a single underlying PFC-mediated cognitive control mechanism and with physiologic dysfunction of the dorsolateral PFC in schizophrenic patients reflecting impairments in this mechanism.     

Further evidence of intact working memory in autism

Earlier investigations have found mixed evidence of working memory impairment in autism. The present study examined working memory in a high-functioning autistic sample, relative to both a clinical control group diagnosed with Tourette Syndrome and a typically developing control group. No group differences were found across three tasks and five dependent measures of working memory. Performance was significantly correlated with both age and IQ. It is concluded that working memory is not one of the executive functions that is seriously impaired in autism. We also suggest that the format of administration of working memory tasks may be important in determining whether or not performance falls in the impaired range.     

Visual working memory as visual attention sustained internally over time

Visual working memory and visual attention are intimately related, such that working memory encoding and maintenance reflects actively sustained attention to a limited number of visual objects and events important for ongoing cognition and action. Although attention is typically considered to operate over perceptual input, a recent taxonomy proposes to additionally consider how attention can be directed to internal perceptual representations in the absence of sensory input, as well as other internal memories, choices, and thoughts (Chun, Golomb, & Turk-Browne, 2011). Such internal attention enables prolonged binding of features into integrated objects, along with enhancement of relevant sensory mechanisms. These processes are all limited in capacity, although different types of working memory and attention, such as spatial vs. object processing, operate independently with separate capacity. Overall, the success of maintenance depends on the ability to inhibit both external (perceptual) and internal (cognitive) distraction. Working memory is the interface by which attentional mechanisms select and actively maintain relevant perceptual information from the external world as internal representations within the mind.     

The contributions of cerebro-cerebellar circuitry to executive verbal working memory

Contributions of cerebro-cerebellar function to executive verbal working memory were examined using event-related functional magnetic resonance imaging (fMRI) while 16 subjects completed two versions of the Sternberg task. In both versions subjects were presented with two or six target letters during the encoding phase, which were held in memory during the maintenance phase. A single probe letter was presented during the retrieval phase. In the “match condition”, subjects decided whether the probe matched the target letters. In the “executive condition”, subjects created a new probe by counting two alphabetical letters forward (e.g., f → h) and decided whether the new probe matched the target letters. Neural activity during the match and executive conditions was compared during each phase of the task. There were four main findings. First, cerebro-cerebellar activity increased as a function of executive load. Second, the dorsal cerebellar dentate co-activated with the supplementary motor area (SMA) during encoding. This likely represented the formation of an articulatory (motor) trajectory. Third, the ventral cerebellar dentate co-activated with anterior prefrontal regions Brodmann Area (BA) 9/46 and the pre-SMA during retrieval. This likely represented the manipulation of information and formation of a response. A functional dissociation between the dorsal “motor” dentate and “cognitive” ventral dentate agrees with neuroanatomical tract tracing studies that have demonstrated separate neural pathways involving each region of the dentate: the dorsal dentate projects to frontal motor areas (including the SMA), and the ventral dentate projects to frontal cognitive areas (including BA 9/46 and the pre-SMA). Finally, activity during the maintenance phase in BA 9, anterior insula, pre-SMA and ventral dentate predicted subsequent accuracy of response to the probe during the retrieval phase. This finding underscored the significant contribution of the pre-SMA/ventral dentate pathway - observed several seconds prior to any motor response to the probe - to executive verbal working memory.     

Role of the left amygdala and right orbital frontal cortex in emotional interference resolution facilitation in working memory

Previous research has shown that emotional information aids conflict resolution in working memory (Levens & Phelps, 2008). Using a Recency-probes working memory (WM) paradigm, Levens and Phelps found that positive and negative emotional stimuli reduced the amount of interference created when information that was once relevant conflicted with currently relevant information, suggesting that emotional information facilitates interference resolution in WM. To determine what regions of the prefrontal cortex (PFC) and temporal lobes are critical to the influence of emotional stimuli on interference resolution, we conducted a Recency-probes emotion paradigm with right and left unilateral frontal and temporal lobe lesion patients. The frontal lobe lesion patient group comprised individuals with unilateral ventral and dorsal PFC lesions. The temporal lobe lesion patient group comprised individuals with lesions of the amygdala and surrounding structures. Results indicate that when the left amygdala is damaged, emotional facilitation of interference resolution is absent (equal emotional and neutral interference levels), when the right orbital frontal cortex (OFC) is damaged, in contrast, emotional interference resolution is impaired (emotional interference levels are higher than neutral levels are). Based on these unique patterns we propose specific contributions for these regions in the emotional facilitation of interference resolution in WM.     

A functional magnetic resonance imaging study of working memory abnormalities in schizophrenia.

BACKGROUND: Previous neuroimaging studies of working memory (WM) in schizophrenia, typically focusing on dorsolateral prefrontal cortex, yield conflicting results, possibly because of varied choice of tasks and analysis techniques. We examined neural function changes at several WM loads to derive a more complete picture of WM dysfunction in schizophrenia. METHODS: We used a version of the Sternberg Item Recognition Paradigm to test WM function at five distinct loads. Eighteen schizophrenia patients and 18 matched healthy controls were scanned with functional magnetic resonance imaging at 3 Tesla. RESULTS: Patterns of both overactivation and underactivation in patients were observed depending on WM load. Patients' activation was generally less responsive to load changes than control subjects', and different patterns of between-group differences were observed for memory encoding and retrieval. In the specific case of successful retrieval, patients recruited additional neural circuits unused by control subjects. Behavioral effects were generally consistent with these imaging results. CONCLUSIONS: Differential findings of overactivation and underactivation may be attributable to patients' decreased ability to focus and allocate neural resources at task-appropriate levels. Additionally, differences between encoding and retrieval suggest that WM dysfunction may be manifested differently during the distinct phases of encoding, maintenance, and retrieval.     

Sustained and transient neural modulations in prefrontal cortex related to declarative long-term memory, working memory, and attention

Common activations in prefrontal cortex (PFC) during episodic and semantic long-term memory (LTM) tasks have been hypothesized to reflect functional overlap in terms of working memory (WM) and cognitive control. To evaluate a WM account of LTM-general activations, the present study took into consideration that cognitive task performance depends on the dynamic operation of multiple component processes, some of which are stimulus-synchronous and transient in nature; and some that are engaged throughout a task in a sustained fashion. PFC and WM may be implicated in both of these temporally independent components. To elucidate these possibilities we employed mixed blocked/event-related functional magnetic resonance imaging (fMRI) procedures to assess the extent to which sustained or transient activation patterns overlapped across tasks indexing episodic and semantic LTM, attention (ATT), and WM. Within PFC, ventrolateral and medial areas exhibited sustained activity across all tasks, whereas more anterior regions including right frontopolar cortex were commonly engaged in sustained processing during the three memory tasks. These findings do not support a WM account of sustained frontal responses during LTM tasks, but instead suggest that the pattern that was common to all tasks reflects general attentional set/vigilance, and that the shared WM-LTM pattern mediates control processes related to upholding task set. Transient responses during the three memory tasks were assessed relative to ATT to isolate item-specific mnemonic processes and were found to be largely distinct from sustained effects. Task-specific effects were observed for each memory task. In addition, a common item response for all memory tasks involved left dorsolateral PFC (DLPFC). The latter response might be seen as reflecting WM processes during LTM retrieval. Thus, our findings suggest that a WM account of shared PFC recruitment in LTM tasks holds for common transient item-related responses rather than sustained state-related responses that are better seen as reflecting more general attentional/control processes.     

Using executive heterogeneity to explore the nature of working memory deficits in Parkinson's disease

Idiopathic Parkinson's disease (IPD) is characterised by a triad of motor symptoms, namely bradykinesia, rigidity and resting tremor, although cognitive impairment is a common feature of the disease and has been accepted as one of the strong predictors of quality of life in such patients. Neuropsychological testing in Parkinson's disease often reveals a pattern of cognitive impairment similar to that observed in patients with frontal lobe lesions, although clear differences between the two groups have also often been reported. This apparent inconsistency in the literature may reflect heterogeneity among different groups of patients with Parkinson's disease, although to date this possibility has not been formally addressed. In this study, two groups of patients with Parkinson's disease were assessed on a novel verbal memory task, which allowed different aspects of working memory function such as maintenance, retrieval and manipulation to be tested within the same general paradigm. The two groups were selected according to either good or bad performance on a 'standard', visuospatial test of executive function (The Tower of London Task), but were well matched on all other demographic and cognitive measures tested. The sub-group of Parkinson's disease patients with Tower of London defined executive deficit were specifically impaired at manipulating information within verbal working memory, both compared to controls and to the group of patients with no predefined executive impairments. In contrast, the three groups did not differ in their ability to maintain or retrieve information within verbal working memory. Given the known preferential role of the dorsolateral frontal cortex in higher executive functions, (including both planning and the manipulation of information within working memory), these results suggest that, in a subset of Parkinson's disease patients, it is the frontostriatal circuitry involving this region which is primarily affected, while other components of this circuitry may be relatively spared. The results also suggest that performance on the Tower of London task may prove to be a useful discriminant variable in defining the nature of cognitive heterogeneity in Parkinson's disease.     

Differences in executive abilities rather than associative processes contribute to memory development

Children''s learning capabilities change while growing up. One framework that describes the cognitive and neural development of children''s growing learning abilities is the two‐component model. It distinguishes processes that integrate separate features into a coherent memory representation (associative component) and executive abilities, such as elaboration, evaluation, and monitoring, that support memory processing (strategic component). In an fMRI study using an object‐location association paradigm, we investigated how the two components influence memory performance across development. We tested children (10–12 years, n = 31), late adolescents (18 years, n = 29), and adults (25+ years, n = 30). For studying the associative component, we also probed how the utilisation of prior knowledge (schemas) facilitates memory across age groups. Children had overall lower retrieval performance, while adolescents and adults did not differ from each other. All groups benefitted from schemas, but this effect did not differ between groups. Performance differences between groups were associated with deactivation of the dorsal medial prefrontal cortex (dmPFC), which in turn was linked to executive functioning. These patterns were stronger in adolescents and adults and seemed absent in children. Thus, the children''s executive system, the strategic component, is not as mature and thus cannot facilitate memory performance in the same way as in adolescents/adults. In contrast, we did not find age‐related differences in the associative component; with activity in the angular gyrus predicting memory performance systematically across groups. Overall, our results suggest that differences of executive rather than associative abilities explain memory differences between children, adolescents, and adults.     

Working memory in Down syndrome: is there a dual task deficit?

Background Recent studies have shown that individuals with Down syndrome (DS) are poorer than controls in performing verbal and visuospatial dual tasks. The present study aims at better investigating the dual task deficit in working memory in individuals with DS. Method Forty‐five individuals with DS and 45 typically developing children matched for verbal mental age completed a series of verbal and visuospatial working memory tasks, involving conditions that either required the combination of two tasks in the same modality (verbal or visual) or of cross‐modality pairs of tasks. Results and conclusions Two distinct deficits were found in individuals with DS: impairment in verbal tasks and further impairment in all dual task conditions. The results confirm the hypothesis of a central executive impairment in individuals with DS.     

Evidence for quantitative domain dominance for verbal and spatial working memory in frontal and parietal cortex

Neuroimaging studies in humans have shown that different working memory (WM) tasks recruit a common bilateral fronto-parietal cortical network. Animal studies as well as neuroimaging studies in humans have suggested that this network, in particular the prefrontal cortex, is preferentially recruited when material from different domains (e.g. spatial information or verbal/object information) has to be memorized. Early imaging studies have suggested qualitative dissociations in the prefrontal cortex for spatial and object/verbal WM, either in a left-right or a ventral-dorsal dimension. However, results from different studies are inconsistent. Moreover, recent fMRI studies have failed to find evidence for domain dependent dissociations of WM-related activity in prefrontal cortex. Here we present evidence from two independent fMRI studies using physically identical stimuli in a verbal and spatial WM task showing that domain dominance for WM does indeed exist, although only in the form of quantitative differences in activation and not in the form of a dissociation with different prefrontal regions showing mutually exclusive activation in different domains. Our results support a mixed dimension model of domain dominance for WM within the prefrontal cortex, with left ventral prefrontal cortex (PFC) supporting preferentially verbal WM and right dorsal PFC supporting preferentially spatial WM. The concept of domain dominance is discussed in the light of recent theories of prefrontal cortex function.