Tracelink: A model of consolidation and amnesia

Meeter, Martijn; Murre, Jaap M. J.

doi:10.1080/02643290442000194

Cited by 48 publications

(67 citation statements)

References 99 publications

(112 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These results suggest a disengagement of the hippocampus from memory retrieval over time and replicate previous findings on system-level consolidation, using a similar face-place task, but without regularities (Takashima et al, 2009). A time-limited involvement of the hippocampus in event memory is in line with a large body of evidence, including studies in animals (Frankland et al, 2004;Maviel et al, 2004;Mishkin, 1978;Quillfeldt et al, 1996;Takehara et al, 2002;Zola-Morgan and Squire, 1985), humans (Bayley et al, 2006;Piefke, 2003;Scoville and Milner, 1957;Teng and Squire, 1999) and in silico (McClelland and Goddard, 1996;McClelland et al, 1995;Meeter and Murre 2005;Murre et al 2006).…”

Section: Discussionsupporting

confidence: 78%

“…A 'direct' route, as suggested above, may therefore, at least to some extent, act in addition to a remaining hippocampal connection. Accordingly, neural network models of system-level consolidation involve an intermediate consolidation state, in which cortico-cortical connectivity in neocortex is building up, while connections between the hippocampus and neocortex still bind the representation (McClelland and Goddard, 1996;McClelland et al, 1995;Meeter and Murre, 2005;Murre et al, 2006).…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Neural mechanisms supporting the extraction of general knowledge across episodic memories

et al. 2014

View full text Add to dashboard Cite

a b s t r a c t a r t i c l e i n f oGeneral knowledge acquisition entails the extraction of statistical regularities from the environment. At high levels of complexity, this may involve the extraction, and consolidation, of associative regularities across event memories. The underlying neural mechanisms would likely involve a hippocampo-neocortical dialog, as proposed previously for system-level consolidation. To test these hypotheses, we assessed possible differences in consolidation between associative memories containing cross-episodic regularities and unique associative memories. Subjects learned face-location associations, half of which responded to complex regularities regarding the combination of facial features and locations, whereas the other half did not. Importantly, regularities could only be extracted over hippocampus-encoded, associative aspects of the items. Memory was assessed both immediately after encoding and 48 h later, under fMRI acquisition. Our results suggest that processes related to system-level reorganization occur preferentially for regular associations across episodes. Moreover, the buildup of general knowledge regarding regular associations appears to involve the coordinated activity of the hippocampus and mediofrontal regions. The putative cross-talk between these two regions might support a mechanism for regularity extraction. These findings suggest that the consolidation of cross-episodic regularities may be a key mechanism underlying general knowledge acquisition.

show abstract

Section: Discussionsupporting

confidence: 78%

Section: Discussionmentioning

confidence: 99%

Neural mechanisms supporting the extraction of general knowledge across episodic memories

et al. 2014

View full text Add to dashboard Cite

show abstract

“…Therefore the ability to learn new items fades gradually without affecting remote memories. The same effect may explain Ribot gradients in retrograde amnesia much better than previous models relying on gradients in replay time [4].Another salient feature of memory is the spacing effect [5] where learning new items is more effective if rehearsal is spread out over time compared to the case when rehearsal is done in a single time block. For example, rehearsing a list of vocabulary two times for ten minutes each is more effective than doing a single rehearsal for twenty minutes.…”

mentioning

confidence: 63%

Structural plasticity, cortical memory, and the spacing effect

Knoblauch

2009

BMC Neurosci

View full text Add to dashboard Cite

The neurophysiological basis of learning and memory is commonly attributed to the modification of synaptic strengths in neuronal networks. Recent experiments suggest also a major role of structural plasticity, including elimination and regeneration of synapses, growth and retraction of dendritic spines, and remodeling of axons and dendrites [1]. Here, I develop a simple model of structural plasticity and synaptic consolidation in neural networks and apply it to Willshaw-type network models of distributed associative memory [2]. The model assumes synapses with discrete weight states. Synapses with low weights have a high probability of being erased and replaced by novel synapses at other locations. In contrast, synapses with large weights are consolidated and cannot be erased. Analysis and numerical simulations reveal that this model can explain various cognitive phenomena much better than alternative network models employing synaptic plasticity only.In previous work, I have shown that networks with low anatomical connectivity employing structural plasticity in coordination with stimulus repetition (e.g., by hippocampal memory replay) can store much more information per synapse by "emulating" high effective memory connectivity close to potential network connectivity [2]. In this work, I present additional simulations and analyses suggesting that networks employing structural plasticity suffer to a much lesser degree from catastrophic forgetting than models without structural plasticity [3] if the number of consolidated synapses remains sufficiently low. The reason for this effect is that early memories get stored with a higher effective connectivity than recent memories. Therefore the ability to learn new items fades gradually without affecting remote memories. The same effect may explain Ribot gradients in retrograde amnesia much better than previous models relying on gradients in replay time [4].Another salient feature of memory is the spacing effect [5] where learning new items is more effective if rehearsal is spread out over time compared to the case when rehearsal is done in a single time block. For example, rehearsing a list of vocabulary two times for ten minutes each is more effective than doing a single rehearsal for twenty minutes. The spacing effect is a remarkably robust phenomenon and can be found in many explicit and implicit memory tasks in humans and many animals being effective over many time scales from single days to months. This suggests that there should be a very general learning mechanism underlying the spacing effect. However, current explanations rather suggest very specific mechanisms referring to detailed implementations of memory systems including attention, novelty, and context processing. While these explanations may well account for some detailed characteristics of the spacing effect in some specific memory tasks, I suggest that the common underlying mechanism at the cellular level is structural plasticity in sparsely connected neural networks. This is supported by simulation exper...

show abstract

“…Activity-dependent Hebbian models allow for studying memory acquisition, showing such effects as faster forgetting of more recent memories. The memory recency effect, known in the psychological literature on memory as the 'Ribbot gradient', has been noticed a long time ago in retrograde amnesia (63)(64)(65), and has also been observed in Alzheimer's patients. Temporal gradients of memory decline and several other experimental phenomena characterizing memory degradation in AD patients have been recreated in Hebbian models.…”

Section: Alzheimer Diseasementioning

confidence: 99%

“…Human memory involves interactions between hippocampal formation, neocortex and neuromodulatory systems, regulating plasticity of synapses depending on the emotional contents of the situation (63,76,83). Such models have been initially created only at the conceptual level, but computational simulations followed (64,65). More realistic memory models that would allow studying the influence of different neurotransmitters on the inter-module inhibition and betweenmodule excitation should help to evaluate potential benefits of new drugs.…”

Section: Alzheimer Diseasementioning

confidence: 99%

Computational Models of Dementia and Neurological Problems

Duch

2007

Methods in Molecular Biology™

View full text Add to dashboard Cite

The final goal of neuroscience is to fully understand neural processes, their relations to mental processes and to cognitive, affective, and behavioral disorders. Computational modeling, although still in its infancy, already plays a central role in this endeavor. A review of different aspects of computational models that help to explain many features of neuropsychological syndromes and psychiatric disease is presented. Recent advances in computational modeling of epilepsy, cortical reorganization after lesions, Parkinson's and Alzheimer diseases are reviewed. Some trends in computational models of brain functions are identified.Key words: neural networks; cognitive computational neuroscience; associative memory models, cortical reorganization, computational models in psychiatry, Parkinson disease, Alzheimer disease. A bit of historyNeuroinformatics has two large branches. On the one hand it provides tools for storage and analysis of information generated by neuroscience. On the other hand it provides simulations and models that capture some aspects of information processing in the brain. Complexity of the brain dynamics may be too high to understand brain's functions in details in conceptual terms. Computational models based on correct principles may capture progressively larger number of essential features of brain dynamics, eventually leading to models of the whole brain that no individual expert will ever be able to understand in details. This situation is analogous to the cell biology, where the sheer number of biomolecules and their interactions will prevent experts from understanding all genetic and metabolic mechanism of a living cell. From the engineering perspective understanding a complex system implies the ability to build a model that behaves in important aspects in the same way as the system that is being modeled. At this point in history computer simulations are the easiest way to build complex models, but progress in building neuromorphic devices that implement some neural functions in hardware may change this situation in future (1).In 1986 two volumes "Parallel Distributed Processing: Explorations in the Microstructure of Cognition" (2), written mostly by psychologist, were published. The first volume of the PDP book (as it was commonly called) focused on general properties of parallel information processing, drawing analogies with human information processing. The Parallel Distributed Processing (PDP) name did not gain popularity, replaced by "connectionism", the name that stressed the importance of connections that pass information between network of nodes representing neural cell assemblies, increasing or inhibiting their activations. One of the chapters described now famous "backpropagation of errors" algorithm that can be used to train a network of simple artificial neurons with a one-way (feedforward) signal flow desired responses to incoming signals. Such artificial neural networks became very useful for medical diagnostics support, signal and image analysis and monitoring, searc...

show abstract

Tracelink: A model of consolidation and amnesia

Cited by 48 publications

References 99 publications

Neural mechanisms supporting the extraction of general knowledge across episodic memories

Neural mechanisms supporting the extraction of general knowledge across episodic memories

Structural plasticity, cortical memory, and the spacing effect

Computational Models of Dementia and Neurological Problems

Contact Info

Product

Resources

About