Neural coding for the retrieval of multiple memory patterns

We investigate the retrieval dynamics in a feature-based semantic memory model, in which the features are coded by neurons of the Hindmarsh-Rose type in the chaotic regime. We consider the retrieval process as consisting of the synchronized firing activity of the neurons coding for the same memory pattern. The retrieval dynamics is investigated for multiple patterns, with particular attention to the case of overlapping memories. In this case, we hypothesize a dynamical nontransitive mechanism based on synchronization, that allows for a shared feature to participate in multiple memory representations. The problem of the choice of a cognitive plausible time-scale for the retrieval analysis is investigated by analyzing the information that can be inferred from finite-time analyses. Different types of indicators are proposed in order to evaluate the temporal dynamics of the neurons engaged in the retrieval process. We interpret the simulation results as suggestive of a role for chaotic dynamics in allowing for flexible composition of elementary meaningful units in memory representations.     

Neural coding for the retrieval of multiple memory patterns

We investigate the retrieval dynamics in a feature-based semantic memory model, in which the features are coded by neurons of the Hindmarsh–Rose type in the chaotic regime. We consider the retrieval process as consisting of the synchronized firing activity of the neurons coding for the same memory pattern. The retrieval dynamics is investigated for multiple patterns, with particular attention to the case of overlapping memories. In this case, we hypothesize a dynamical nontransitive mechanism based on synchronization, that allows for a shared feature to participate in multiple memory representations. The problem of the choice of a cognitive plausible time-scale for the retrieval analysis is investigated by analyzing the information that can be inferred from finite-time analyses. Different types of indicators are proposed in order to evaluate the temporal dynamics of the neurons engaged in the retrieval process. We interpret the simulation results as suggestive of a role for chaotic dynamics in allowing for flexible composition of elementary meaningful units in memory representations.     

Extremely Dilute Modular Neuronal Networks: Neocortical Memory Retrieval Dynamics

A model of columnar networks of neocortical association areas is studied. The neuronal network is composed of many Hebbian autoassociators, or modules, each of which interacts with a relatively small number of the others, randomly chosen. Any module encodes and stores a number of elementary percepts, or features. Memory items, or patterns, are peculiar combinations of features sparsely distributed over the multi-modular network. Any feature stored in any module can be involved in several of the stored patterns; feature-sharing is in fact source of local ambiguities and, consequently, a potential cause of erroneous memory retrieval spreading through the model network in pattern completion tasks.The memory retrieval dynamics of the large modular autoassociator is investigated by combining mathematical analysis and numerical simulations. An oscillatory retrieval process is proposed that is very efficient in overcoming feature-sharing drawbacks; it requires a mechanism that modulates the robustness of local attractors to noise, and neuronal activity sparseness such that quiescent and active modules are about equally noisy to any post-synaptic module.Moreover, it is shown that statistical correlation between 'kinds' of features across the set of memory patterns can be exploited to obtain a more efficient achievement of memory retrieval capabilities.It is also shown that some spots of the network cannot be reached by retrieval activity spread if they are not directly cued by the stimulus. The locations of these activity isles depend on the pattern to retrieve, while their extension only depends (in large networks) on statistics of inter-modular connections and stored patterns. The existence of activity isles determines an upper-bound to retrieval quality that does not depend on the specific retrieval dynamics adopted, nor on whether feature-sharing is permitted. The oscillatory retrieval process nearly saturates this bound.     

Functional neuroanatomy of the autobiographical memory

Autobiographical memory refers to events and information about personal life and the self. Within autobiographical memory, many authors make a difference between episodic and semantic components. Study of retrograde amnesia gives information about memory consolidation. According to the "standard model" of consolidation, the medial temporal lobe plays a time-limited role in retrieval memory. Functional neuroanatomy studies of autobiographical memory are very few and many are recent. These studies concern which brain regions are involved in the autobiographical retrieval, episodic or semantic autobiographical memory and consolidation process. Results show that autobiographical retrieval depends on specific brain regions like frontal cortex. Concerning memory consolidation, findings are most consistent with the idea that hippocampal complex is involved in both recent and remote memories.     

Synchronization of neural activity is a promising mechanism of memory information processing in networks of columns

Synchronization of the oscillatory discharge of cortical neurons could be a part of the mechanism that is involved in cortical information processing. On the assumption that the basic functional unit is the column composed of local excitatory and inhibitory cells and generating oscillatory neural activity, a network model that attains associative memory function is proposed. The synchronization of oscillation in the model is studied analytically using a sublattice analysis. In particular, the retrieval of a single memory pattern can be studied in the system, which can be derived from the original network model of interacting columns and is formally equivalent to a system of an isolated column. The network model simulated numerically shows a remarkable performance in which retrieval is achieved simultaneously for more than one memory pattern. The manifestations of this simultaneous retrieval in the network dynamics are successive transitions of the network state from a synchronized oscillation for a memory pattern to that for another memory pattern.     

Recognition memory for vibrotactile rhythms: an fMRI study in blind and sighted individuals

Calcarine sulcal cortex possibly contributes to semantic recognition memory in early blind (EB). We assessed a recognition memory role using vibrotactile rhythms and a retrieval success paradigm involving learned "old" and "new" rhythms in EB and sighted. EB showed no activation differences in occipital cortex indicating retrieval success but replicated findings of somatosensory processing. Both groups showed retrieval success in primary somatosensory, precuneus, and orbitofrontal cortex. The S1 activity might indicate generic sensory memory processes.     

Task and content modulate amygdala-hippocampal connectivity in emotional retrieval

The ability to remember emotional events is crucial for adapting to biologically and socially significant situations. Little is known, however, about the nature of the neural interactions supporting the integration of mnemonic and emotional information. Using fMRI and dynamic models of effective connectivity, we examined regional neural activity and specific interactions between brain regions during a contextual memory retrieval task. We independently manipulated emotional context and relevance of retrieved emotional information to task demands. We show that retrieval of emotionally valenced contextual information is associated with enhanced connectivity from hippocampus to amygdala, structures crucially involved with encoding of emotional events. When retrieval of emotional information is relevant to current behavior, amygdala-hippocampal connectivity increases bidirectionally, under modulatory influences from orbitofrontal cortex, a region implicated in representation of affective value and behavioral guidance. Our findings demonstrate that both memory content and behavioral context impact upon large scale neuronal dynamics underlying emotional retrieval.     

Directional Theta Coherence in Prefrontal Cortical to Amygdalo-Hippocampal Pathways Signals Fear Extinction

Theta oscillations are considered crucial mechanisms in neuronal communication across brain areas, required for consolidation and retrieval of fear memories. One form of inhibitory learning allowing adaptive control of fear memory is extinction, a deficit of which leads to maladaptive fear expression potentially leading to anxiety disorders. Behavioral responses after extinction training are thought to reflect a balance of recall from extinction memory and initial fear memory traces. Therefore, we hypothesized that the initial fear memory circuits impact behavioral fear after extinction, and more specifically, that the dynamics of theta synchrony in these pathways signal the individual fear response. Simultaneous multi-channel local field and unit recordings were obtained from the infralimbic prefrontal cortex, the hippocampal CA1 and the lateral amygdala in mice. Data revealed that the pattern of theta coherence and directionality within and across regions correlated with individual behavioral responses. Upon conditioned freezing, units were phase-locked to synchronized theta oscillations in these pathways, characterizing states of fear memory retrieval. When the conditioned stimulus evoked no fear during extinction recall, theta interactions were directional with prefrontal cortical spike firing leading hippocampal and amygdalar theta oscillations. These results indicate that the directional dynamics of theta-entrained activity across these areas guide changes in appraisal of threatening stimuli during fear memory and extinction retrieval. Given that exposure therapy involves procedures and pathways similar to those during extinction of conditioned fear, one therapeutical extension might be useful that imposes artificial theta activity to prefrontal cortical-amygdalo-hippocampal pathways that mimics the directionality signaling successful extinction recall.     

Competitive Semantic Memory Retrieval: Temporal Dynamics Revealed by Event-Related Potentials

Memories compete for retrieval when they are related to a common retrieval cue. Previous research has shown that retrieval of a target memory may lead to subsequent retrieval-induced forgetting (RIF) of currently irrelevant competing memories. In the present study, we investigated the time course of competitive semantic retrieval and examined the neurocognitive mechanisms underlying RIF. We contrasted two theoretical accounts of RIF by examining a critical aspect of this memory phenomenon, namely the extent to which it depends on successful retrieval of the target memory. Participants first studied category-exemplar word-pairs (e.g. Fruit—Apple). Next, we recorded electrophysiological measures of brain activity while the participants performed a competitive semantic cued-recall task. In this task, the participants were provided with the studied categories but they were instructed to retrieve other unstudied exemplars (e.g. Fruit—Ma__?). We investigated the event-related potential (ERP) correlates of retrieval success by comparing ERPs from successful and failed retrieval trials. To isolate the ERP correlates of continuous retrieval attempts from the ERP correlates of retrieval success, we included an impossible retrieval condition, with incompletable word-stem cues (Drinks—Wy__) and compared it with a non-retrieval presentation baseline condition (Occupation—Dentist). The participants’ memory for all the studied exemplars was tested in the final phase of the experiment. Taken together, the behavioural results suggest that RIF is independent of target retrieval. Beyond investigating the mechanisms underlying RIF, the present study also elucidates the temporal dynamics of semantic cued-recall by isolating the ERP correlates of retrieval attempt and retrieval success. The ERP results revealed that retrieval attempt is reflected in a late posterior negativity, possibly indicating construction of candidates for completing the word-stem cue and retrieval monitoring whereas retrieval success was reflected in an anterior positive slow wave.     

Effects of Hippocampal LIMK Inhibition on Memory Acquisition,Consolidation, Retrieval,Reconsolidation, and Extinction

Long-lasting changes in dendritic spines provide a physical correlate for memory formation and persistence. LIM kinase (LIMK) plays a critical role in orchestrating dendritic actin dynamics during memory processing, since it is the convergent downstream target of both the Rac1/PAK and RhoA/ROCK pathways that in turn induce cofilin phosphorylation and prevent depolymerization of actin filaments. Here, using a potent LIMK inhibitor (BMS-5), we investigated the role of LIMK activity in the dorsal hippocampus during contextual fear memory in rats. We first found that post-training administration of BMS-5 impaired memory consolidation in a dose-dependent manner. Inhibiting LIMK before training also disrupted memory acquisition. We then demonstrated that hippocampal LIMK activity seems to be critical for memory retrieval and reconsolidation, since both processes were impaired by BMS-5 treatment. Contextual fear memory extinction, however, was not sensitive to the same treatment. In conclusion, our findings demonstrate that hippocampal LIMK activity plays an important role in memory acquisition, consolidation, retrieval, and reconsolidation during contextual fear conditioning.     

Investigation of changes in EEG complexity during memory retrieval: the effect of midazolam

The aim of this study is applying nonlinear methods to assess changes in brain dynamics in a placebo-controlled study of midazolam-induced amnesia. Subjects injected with saline and midazolam during study, performed old/new recognition memory tests with EEG recording. Based on previous studies, as midazolam causes anterograde amnesia, we expected that midazolam would affect the EEG’s degree of complexity. Recurrence quantification analysis, and approximate entropy were used in this assessment. These methods compare with other nonlinear techniques such as computation of the correlation dimension, are suitable for non-stationary EEG signals. Our findings suggest that EEG’s complexity decreases during memory retrieval. Although this trend is observed in nonlinear curves related to the midazolam condition, the overall complexity were greater than in the saline condition. This result implies that impaired memory function caused by midazolam is associated with greater EEG’s complexity compared to normal memory retrieval in saline injection.     

Interaction of N-methyl-D-aspartate and Dopamine D1 receptors in modulation of passive avoidance extinction at mice with depressive-like state

The effect of activation of N-methyl-D-aspartate (D-cycloserine) and dopamine D1 (SKF 38393) receptors on learning and extinction of the passive avoidance response in mice under normal conditions and after formation of "behavioural despair" is studied. The data on ineffectiveness of D-cycloserine and SKF 38393 on training a conditional reflex were obtained. In mice with the normal state, SKF 38393 did not alter the dynamics of extinction, and D-cycloserine facilitated a more rapid decline in retrieval of memory trace when testing without penalty. On exposure to D-cycloserine + SKF 38393 injection, dynamics of extinction was similar to that under the action of D-cycloserine. In mice with the reaction of "behavioral despair", D-cycloserine and SKF 38393 reduced the deficit of the passive avoidance extinction typical for "depressed" animals without drugs. With simultaneous activation of NMDA and D1 receptors we observed acceleration of the extinction start and development of complete extinction of the memory trace about pain impact as compared with single injections of D-cycloserine and SKF 38393.     

Sleep loss produces false memories

People sometimes claim with high confidence to remember events that in fact never happened, typically due to strong semantic associations with actually encoded events. Sleep is known to provide optimal neurobiological conditions for consolidation of memories for long-term storage, whereas sleep deprivation acutely impairs retrieval of stored memories. Here, focusing on the role of sleep-related memory processes, we tested whether false memories can be created (a) as enduring memory representations due to a consolidation-associated reorganization of new memory representations during post-learning sleep and/or (b) as an acute retrieval-related phenomenon induced by sleep deprivation at memory testing. According to the Deese, Roediger, McDermott (DRM) false memory paradigm, subjects learned lists of semantically associated words (e.g., "night", "dark", "coal",...), lacking the strongest common associate or theme word (here: "black"). Subjects either slept or stayed awake immediately after learning, and they were either sleep deprived or not at recognition testing 9, 33, or 44 hours after learning. Sleep deprivation at retrieval, but not sleep following learning, critically enhanced false memories of theme words. This effect was abolished by caffeine administration prior to retrieval, indicating that adenosinergic mechanisms can contribute to the generation of false memories associated with sleep loss.     

High Working Memory Capacity Predicts Less Retrieval Induced Forgetting

Background

Working Memory Capacity (WMC) is thought to be related to executive control and focused memory search abilities. These two hypotheses make contrasting predictions regarding the effects of retrieval on forgetting. Executive control during memory retrieval is believed to lead to retrieval induced forgetting (RIFO) because inhibition of competing memory traces during retrieval renders them temporarily less accessible. According to this suggestion, superior executive control should increase RIFO. Alternatively, superior focused search abilities could diminish RIFO, because delimiting the search set reduces the amount of competition between traces and thus the need for inhibition. Some evidence suggests that high WMC is related to more RIFO, which is inconsistent with the focused search hypothesis.

Methodology/Principal Findings

Using the RIFO paradigm, we created distinct and overlapping categories to manipulate the amount of competition between them. This overlap increased competition between some categories while exclusive use of weak exemplars ensured negligible effects of output interference and integration.Low WMC individuals exhibited RIFO within and between overlapping categories, indicating the effect of resolving competition during retrieval. High WMC individuals only exhibited between-category RIFO, suggesting they experienced reduced competition resolution demands. Low WMC Individuals exhibited the strongest RIFO and no retrieval benefits when interference resolution demands were high.

Conclusions/Significance

Our findings qualify the inhibitory explanation for RIFO by incorporating the focused search hypothesis for materials that are likely to pose extraordinary challenges at retrieval. The results highlight the importance of considering individual differences in retrieval-induced effects and qualify existing models of these effects.     

Dissociating memory retrieval processes using fMRI: evidence that priming does not support recognition memory

We employed event-related fMRI to constrain cognitive accounts of memory retrieval. Studies of explicit retrieval reveal that lateral and medial parietal, dorsal middle frontal gyrus, and anterior prefrontal cortex respond more for studied than new words, reflecting a correlate of "retrieval success." Studies of implicit memory suggest left temporal cortex, ventral and dorsal inferior frontal gyrus respond less for studied than new words, reflecting a correlate of "conceptual priming." In the present study, responses for old and new items were compared during performance on explicit recognition (old/new judgement) and semantic (abstract/concrete judgement) tasks. Regions associated with priming were only modulated during the semantic task, whereas regions associated with retrieval success were modulated during both tasks. These findings constrain functional-anatomic accounts of the networks, suggesting that processes associated with priming do not support explicit recognition judgments.     

Place cells, spatial maps and the population code for memory

The study of population dynamics in hippocampal place cells has emerged as one of the most powerful tools for understanding the encoding, storage and retrieval of declarative memory. Recent work has laid out the contours of an attractor-based hippocampal population code for memory in recurrent circuits of the hippocampus. The code is based on inputs from a topographically organized, path-integration-dependent spatial map that lies upstream in the medial entorhinal cortex. The recurrent networks of the hippocampal formation enable these spatial inputs to be synthesized with nonspatial event-related information.     

Dynamics control of semantic processes in a hierarchical associative memory

A neural mechanism for control of dynamics and function of associative processes in a hierarchical memory system is demonstrated. For the representation and processing of abstract knowledge, the semantic declarative memory system of the human brain is considered. The dynamics control mechanism is based on the influence of neuronal adaptation on the complexity of neural network dynamics. Different dynamical modes correspond to different levels of the ultrametric structure of the hierarchical memory being invoked during an associative process. The mechanism is deterministic but may also underlie free associative thought processes. The formulation of an abstract neural network model of hierarchical associative memory utilizes a recent approach to incorporate neuronal adaptation. It includes a generalized neuronal activation function recently derived by a Hodgkin-Huxley-type model. It is shown that the extent to which a hierarchically organized memory structure is searched is controlled by the neuronal adaptability, i.e. the strength of coupling between neuronal activity and excitability. In the brain, the concentration of various neuromodulators in turn can regulate the adaptability. An autonomously controlled sequence of bifurcations, from an initial exploratory to a final retrieval phase, of an associative process is shown to result from an activity-dependent release of neuromodulators. The dynamics control mechanism may be important in the context of various disorders of the brain and may also extend the range of applications of artificial neural networks.     

An algebraic model of an associative noise-like coding memory

A mathematical model of an associative memory is presented, sharing with the optical holography memory systems the properties which establish an analogy with biological memory. This memory system-developed from Gabor's model of memoryis based on a noise-like coding of the information by which it realizes a distributed, damage-tolerant, equipotential storage through simultaneous state changes of discrete substratum elements. Each two associated items being stored are coded by each other by means of two noise-like patterns obtained from them through a randomizing preprocessing. The algebraic braic transformations operating the information storage and retrieval are matrix-vector products involving Toeplitz type matrices. Several noise-like coded memory traces are superimposed on a common substratum without crosstalk interference; moreover, extraneous noise added to these memory traces does not injure the stored information. The main performances shown by this memory model are: i) the selective, complete recovering of stored information from incomplete keys, both mixed with extraneous information and translated from the position learnt; ii) a dynamic recollection where the information just recovered acts as a new key for a sequential retrieval process; iii) context-dependent responses. The hypothesis that the information is stored in the nervous system through a noise-like coding is suggested. The model has been simulated on a digital computer using bidimensional images.     

Dynamics control of semantic processes in a hierarchical associative memory

 A neural mechanism for control of dynamics and function of associative processes in a hierarchical memory system is demonstrated. For the representation and processing of abstract knowledge, the semantic declarative memory system of the human brain is considered. The dynamics control mechanism is based on the influence of neuronal adaptation on the complexity of neural network dynamics. Different dynamical modes correspond to different levels of the ultrametric structure of the hierarchical memory being invoked during an associative process. The mechanism is deterministic but may also underlie free associative thought processes. The formulation of an abstract neural network model of hierarchical associative memory utilizes a recent approach to incorporate neuronal adaptation. It includes a generalized neuronal activation function recently derived by a Hodgkin-Huxley-type model. It is shown that the extent to which a hierarchically organized memory structure is searched is controlled by the neuronal adaptability, i.e. the strength of coupling between neuronal activity and excitability. In the brain, the concentration of various neuromodulators in turn can regulate the adaptability. An autonomously controlled sequence of bifurcations, from an initial exploratory to a final retrieval phase, of an associative process is shown to result from an activity-dependent release of neuromodulators. The dynamics control mechanism may be important in the context of various disorders of the brain and may also extend the range of applications of artificial neural networks. Received: 19 April 1995/Accepted in revised form: 8 August 1995     

Chaotic EEG patterns during recall of stressful memory related to panic attack

Chaotic transitions likely emerge in a wide variety of cognitive phenomena and may be linked to specific changes during the development of mental disorders. They represent relatively short periods in the behavior of a system, which are extremely sensitive to very small changes. This increased sensitivity has been suggested to occur also during retrieval of stressful emotional experiences because of their fragmentary, temporally and spatially disorganized character. To test this hypothesis we recorded EEG during retrieval of fearful memories related to panic attack in 7 patients and retrieval of anxiety-related memories in 11 healthy controls. Nonlinear data analysis of EEG records showed a statistically significant increase in degree of chaotic dynamics after retrieval of stressful memories in majority of patients as well as in control subjects. This change correlated with subjective intensity of anxiety induced during the memory retrieval. The data suggest a role of nonlinear changes of neural dynamics in the processing of stressful anxiety-related memories, which may play an important role in the pathophysiology of panic disorder.