ArticlePublisher preview available

Neural Activity Reveals Interactions Between Episodic and Semantic Memory Systems During Retrieval

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

Whereas numerous findings support a distinction between episodic and semantic memory, it is now widely acknowledged that these two forms of memory interact during both encoding and retrieval. The precise nature of this interaction, however, remains poorly understood. To examine the role of semantic organization during episodic encoding and retrieval, we recorded intracranial encephalographic signals as 69 neurosurgical patients studied and subsequently recalled categorized and unrelated word lists. Applying multivariate classifiers to neural recordings, we were able to reliably predict encoding success, retrieval success, and temporal and categorical clustering during recall. By assessing how these classifiers generalized across list types, we identified specific retrieval processes that predicted recall of categorized lists and distinguished between recall transitions within and between category clusters. These results particularly implicate retrieval (rather than encoding) processes in the categorical organization of episodic memories.
Neural Activity Reveals Interactions Between Episodic and Semantic
Memory Systems During Retrieval
Christoph T. Weidemann
Swansea University and University of Pennsylvania James E. Kragel
University of Pennsylvania
Bradley C. Lega
University of Texas Southwestern Gregory A. Worrell
Mayo Clinic, Rochester, Minnesota
Michael R. Sperling and Ashwini D. Sharan
Thomas Jefferson University Hospital, Philadelphia,
Pennsylvania
Barbara C. Jobst
Darmouth-Hitchcock Medical Center, Lebanon, New Hampshire
Fatemeh Khadjevand
Mayo Clinic, Rochester, Minnesota Kathryn A. Davis
Hospital of the University of Pennsylvania, Philadelphia,
Pennsylvania
Paul A. Wanda, Allison Kadel, Daniel S. Rizzuto, and Michael J. Kahana
University of Pennsylvania
Whereas numerous findings support a distinction between episodic and semantic memory, it is now
widely acknowledged that these two forms of memory interact during both encoding and retrieval. The
precise nature of this interaction, however, remains poorly understood. To examine the role of semantic
organization during episodic encoding and retrieval, we recorded intracranial encephalographic signals
as 69 neurosurgical patients studied and subsequently recalled categorized and unrelated word lists.
Applying multivariate classifiers to neural recordings, we were able to reliably predict encoding success,
retrieval success, and temporal and categorical clustering during recall. By assessing how these classifiers
generalized across list types, we identified specific retrieval processes that predicted recall of categorized
lists and distinguished between recall transitions within and between category clusters. These results
particularly implicate retrieval (rather than encoding) processes in the categorical organization of
episodic memories.
Keywords: free recall, episodic memory, semantic memory, intracranial EEG, machine learning
Christoph T. Weidemann, Department of Psychology, Swansea Uni-
versity, and Department of Psychology, University of Pennsylvania;
James E. Kragel, Department of Psychology, University of Pennsylva-
nia; Bradley C. Lega, Department of Neurosurgery, University of Texas
Southwestern; Gregory A. Worrell, Department of Neurology, Mayo
Clinic, Rochester, Minnesota; Michael R. Sperling and Ashwini D.
Sharan, Department of Neurology, Thomas Jefferson University Hos-
pital, Philadelphia, Pennsylvania; Barbara C. Jobst, Department of
Neurology, Darmouth-Hitchcock Medical Center, Lebanon, New
Hampshire; Fatemeh Khadjevand, Department of Neurology, Mayo
Clinic; Kathryn A. Davis, Department of Neurology, Hospital of the
University of Pennsylvania, Philadelphia, Pennsylvania; Paul A.
Wanda, Allison Kadel, Daniel S. Rizzuto, and Michael J. Kahana,
Department of Psychology, University of Pennsylvania.
The authors thank Blackrock Microsystems for providing neural
recording equipment. This work was supported by the DARPA Restor-
ing Active Memory (RAM) program (Cooperative Agreement N66001-
14-2-4032) and by the National Institutes of Health (Grant MH055687).
The views, opinions, and/or findings contained in this material are those
of the authors and should not be interpreted as representing the official
views or policies of the Department of Defense or the U.S. Government.
Christoph T. Weidemann & James E. Kragel contributed equally to this
work. Christoph T. Weidemann & James E. Kragel analyzed the data
and wrote the paper; Christoph T. Weidemann, James E. Kragel, &
Michael J. Kahana designed analyses and edited the paper; Bradley C.
Lega, Gregory A. Worrell, Michael R. Sperling, Ashwini D. Sharan,
Barbara C. Jobst, & Kathryn A. Davis recruited participants and pro-
vided general assistance; Fatemeh Khadjevand, Paul A. Wanda, &
Allison Kadel collected data; Daniel S. Rizzuto & Michael J. Kahana
designed experiments. We thank Youssef Ezzyat for insightful discus-
sions. The authors declare that they have no competing financial inter-
ests. Part of this work was presented at the 2018 Context and Episodic
Memory Symposium in Philadelphia, Pennsylvania and at the 2018
Annual meeting of the Psychonomic Society, New Orleans, Louisiana.
All of the de-identified raw data and analysis code used in this study
may be freely downloaded from the Cognitive Electrophysiology Data
Portal (http://memory.psych.upenn.edu/Electrophysiological_Data).
Correspondence concerning this article should be addressed to Christoph
T. Weidemann, Department of Psychology, Swansea University, Singleton
Park, Swansea SA2 8PP, UK. E-mail: ctw@cogsci.info
THIS ARTICLE HAS BEEN CORRECTED. SEE LAST PAGE
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Journal of Experimental Psychology: General
© 2019 American Psychological Association 2019, Vol. 148, No. 1, 1–12
0096-3445/19/$12.00 http://dx.doi.org/10.1037/xge0000480
1
... Whereas the previously cited studies examined the univariate relation between spectral EEG features and memory processes, other studies have taken a multivariate approach, using machine learning classifiers to distinguish single-trial encoding of subsequently remembered and forgotten items. These multivariate studies have found a similar reliance on combined HFA and LFA effects in predicting memory success [12][13][14][15][16] ; two of these studies 13,16 also showed that multivariate classifiers could distinguish moments preceding correct recalls from periods of deliberation that did not lead to successful recall. Non-invasive EEG and MEG studies have also found that greater HFA and diminished alpha-band activity generally accompany successful as compared with unsuccessful memory encoding and retrieval [17][18][19][20][21] ; theta-band activity exhibits increases in some studies and decreases in other studies [17][18][19][20][21][22][23] . ...
... Whereas the previously cited studies examined the univariate relation between spectral EEG features and memory processes, other studies have taken a multivariate approach, using machine learning classifiers to distinguish single-trial encoding of subsequently remembered and forgotten items. These multivariate studies have found a similar reliance on combined HFA and LFA effects in predicting memory success [12][13][14][15][16] ; two of these studies 13,16 also showed that multivariate classifiers could distinguish moments preceding correct recalls from periods of deliberation that did not lead to successful recall. Non-invasive EEG and MEG studies have also found that greater HFA and diminished alpha-band activity generally accompany successful as compared with unsuccessful memory encoding and retrieval [17][18][19][20][21] ; theta-band activity exhibits increases in some studies and decreases in other studies [17][18][19][20][21][22][23] . ...
... Consistent with univariate analyses, these individualparticipant classifiers revealed elevated high-frequency (gamma) activity and suppressed alpha activity as spectral signatures of good memory (Fig. 4). Our memory encoding classification results also align with prior intracranial EEG findings 12,13,16,33 . In the case of retrieval, however, our study demonstrates participant-specific classification of both prior-list and extra-list intrusions (Fig. 5). ...
Article
Full-text available
Encoding- and retrieval-related neural activity jointly determine mnemonic success. We ask whether electroencephalographic activity can reliably predict encoding and retrieval success on individual trials. Each of 98 participants performed a delayed recall task on 576 lists across 24 experimental sessions. Logistic regression classifiers trained on spectral features measured immediately preceding spoken recall of individual words successfully predict whether or not those words belonged to the target list. Classifiers trained on features measured during word encoding also reliably predict whether those words will be subsequently recalled and further predict the temporal and semantic organization of the recalled items. These findings link neural variability predictive of successful memory with item-to-context binding, a key cognitive process thought to underlie episodic memory function.
... For encoding, participants were sequentially presented 12 words centered on the screen that were selected at random-without replacement in each whole session or two consecutive half sessions-from a pool of 300 highfrequency, intermediate-memorable English or Spanish nouns (http:// memory.psych.upenn.edu/WordPools Weidemann et al., 2019). Each word was presented for 1.6 s with a jittered 0.75-1.2 ...
... We perform all analyses in this paper on a variant of the free recall paradigm called categorized free recall, which uses lists composed of words with semantic relationships. For every whole session (or consecutive half sessions), words were drawn from a pool of 300 that included 12 words each from 25 categories created using Amazon Mechanical Turk to crowdsource typical exemplars for each category (Weidemann et al., 2019). For each list, three semantic categories were randomly chosen, and the four words from each category were presented sequentially in pairs (Fig. 1b). ...
... Hippocampal ripples do not exhibit a SME To clarify the relation between ripples and memory encoding, we align hippocampal recordings ( Fig. 1d and f ) to the onset of word presentation during the study phase of a categorized, delayed free recall task (Weidemann et al., 2019;Fig. 1b), in which participants view a list of words and subsequently recall as many as possible after a distractor period. ...
Article
Full-text available
Direct human brain recordings have confirmed the presence of high-frequency oscillatory events, termed ripples, during awake behavior. While many prior studies have focused on medial temporal lobe (MTL) ripples during memory retrieval, here we investigate ripples during memory encoding. Specifically, we ask whether ripples during encoding predict whether and how memories are subsequently recalled. Detecting ripples from MTL elec-trodes implanted in 116 neurosurgical participants (n = 61 male) performing a verbal episodic memory task, we find that encoding ripples do not distinguish recalled from not recalled items in any MTL region, even as high-frequency activity (HFA) during encoding predicts recall in these same regions. Instead, hippocampal ripples increase during encoding of items that subsequently lead to recall of temporally and semantically associated items during retrieval, a phenomenon known as clustering. This subsequent clustering effect arises specifically when hippocampal ripples co-occur during encoding and retrieval, suggesting that ripples mediate both encoding and reinstatement of episodic memories.
... Because seizures often localize to a small region of the brain, the six channels of the RNS have proven adequate for controlling seizures, or at least significantly reducing seizure frequency, in many patients. However, memory-related brain signals occur across a distributed network of brain regions [6,26] and, as such, require the widespread surveillance capabilities of the SNS and the ability to adaptively select among a larger set of stimulation targets. Analyses of large open-data sets [27,28] informed the SNS design by identifying whole-brain patterns of neural activity that signal periods of good and poor memory. ...
Preprint
Electrical stimulation of the human brain has emerged as a powerful therapeutic modality, enabling the alteration of neural circuits underlying cognition and behavior. Recent evidence indicates that stimulation's effects on physiology and behavior depend on endogenous variation in brain state, as measured by field-potential recordings. Here, we describe a 60-channel brain-computer interface — the Smart Neurostimulation System (SNS) — that combines closed-loop analysis of spectral features of the field potential with multi-channel stimulation capabilities. We demonstrate system functionality via bench tests and from an in vivo ovine study exercising a subset of the device's functionality. Our ovine study shows that the SNS can reliably measure neural correlates of behavior (motion) and the physiological effects of stimulation. We demonstrate the safety of stimulation via histology following a 120-day stimulation study.
... Our findings may be attributable to diminished motor activity, which reflects a decreased capacity for the independent manipulation of objects within the patient's daily environment. As we engage with the world, we form semantic associations through interactions with surrounding objects (Weidemann et al., 2019). The lack of such interaction could reduce their capacity to explore and identify associative connections. ...
Article
Full-text available
Introduction The relationship between motor and cognitive skills is a pivotal issue in neuroscience, with embodied cognition theory asserting that bodily actions and experiences play a vital role in cognitive processing. This relevance is particularly noted in children with severe motor disorders (MD) from birth, highlighting a need to explore how these disorders may impede cognitive functions. Methods In this study, we assessed verbal fluency, a critical component of speech function, in 68 children aged 7 to 15. This group consisted of 36 children with motor disorders, specifically those diagnosed with obstetric brachial plexus palsy (OBPP, n = 22) or arthrogryposis multiplex congenita (AMC, n = 14), and 32 healthy control children. We compared levels of verbal fluency, action/verbal naming, and the development of semantic associations between the two groups. Results The findings revealed that children with motor impairments exhibited significantly lower performance in tasks measuring verbal fluency and semantic association compared to the control group. Mainly, MD children produced fewer words during verbal fluency tasks and demonstrated reduced semantic associations. Interestingly, MD children with unilateral limb impairment outperformed those with bilateral impairment on semantic association tasks. Discussion These results suggest that the cognitive deficits observed in children with motor impairments can be attributed to less engagement with their physical environment, which influences their ability to perceive and manipulate objects differently based on their level of impairment. Additionally, the findings underscore how social and cultural contexts may be affected by motor impairments. Overall, our study supports the concept of embodied cognition, demonstrating that delays in motor skill development among children with OBPP and AMC can harm their cognitive functions.
... In our dataset, a relative gradual functional connectivity was evidenced, from semantic to episodic memories, between widespread temporal areas in both mean connectivity at the global network scale or in the number of connections between nodes. Such a gradation resonates with theories that assume a continuum between different memory systems (Barzykowski & Moulin, 2022;Greenberg & Verfaellie, 2010;Renoult et al., 2012Renoult et al., , 2019 and the evidence of interactions between semantic and episodic memory during retrieval (Greve et al., 2007;Weidemann et al., 2019). It matches recent fMRI evidence that demonstrates a gradual magnitude of activation within a common network of brain regions, from general to autobiographical facts, from autobiographical facts to repeated events, and from repeated events to unique events (Tanguay et al., 2023). ...
... Following this word encoding period, a 20 s self-paced math distractor was provided with problems of summing 3 single digit values, followed by a 30 s free recall period. Words were selected from a previously reported wordpool of 300 common English words, with each word presented only once per session (Weidemann et al., 2019). ...
Preprint
Full-text available
Human and animal studies have implicated oscillatory reset following a behaviorally relevant stimulus in learning and memory function. Here we asked whether phase reset predicts the dynamics of memory encoding for subsequently recalled and forgotten items. Our dataset comprised 188 neurosurgical patients (99 male, 89 female) who studied lists of common words for a subsequent delayed free recall test. We observed significant 3-21Hz overall phase reset widespread throughout the brain in the 125-500ms interval following stimulus presentation for both subsequently recalled and forgotten items. We found that the increased recalls of the primacy effect correlated significantly and exclusively with theta phase reset from 3-7Hz throughout the brain. When controlling for serial position effects and examining the phase consistency differences of subsequently recalled and forgotten items, we identified both a theta (3-6Hz) and an alpha (9-14Hz) component of phase reset. Examined separately, early list items revealed a significant theta phase reset difference, while later list items revealed a significant alpha phase reset difference for subsequently recalled items. In regional analyses, the lateral temporal cortex exhibited a significant theta phase reset difference, while the prefrontal cortex revealed both a significant theta and a significant alpha phase reset difference for subsequently recalled items. These findings support the view that oscillatory reset serves an important role in encoding new episodic associations, and reveal differing oscillatory mechanisms across theta and alpha bands for serial position effects of phase reset versus serial position controlled subsequent recall.
... Emotionally salient events demonstrate higher activation of the amygdala and corresponding neural circuits, leading to better memory formation ( Hammack et al., 2023). Semantic memory recruits many of the same neural circuits used in episodic memory ( Tanguay et al., 2023), and the two memory domains interact heavily in memory formation and retrieval ( Weidemann et al., 2019). The lateral occipital cortex has additionally been found to help integrate new episodic memories into existing semantic knowledge ( Guo & Yang, 2023;Karanian & Slotnick, 2015). ...
Article
Full-text available
Tau pathology in Alzheimer disease (AD) is often evaluated in regions associated with episodic memory impairment. However, heterogeneous spreading patterns of tau are observed and correspond to impairment in different cognitive domains. We have previously developed a metric to quantify tau spread extent that is robustly sensitive to atypical spreading patterns. Here, we evaluate tau spread relative to domain-specific and general cognitive impairments during early stages of AD. In total, 529 participants with baseline tau positron emission tomography (PET) and neuropsychological testing were separated into disease-stage groups based on amyloid PET positivity and clinical status via Clinical Dementia Rating® (CDR®). General cognition was assessed using the Knight Preclinical Alzheimer Cognitive Composite (Knight PACC). Domain-specific composites were calculated for episodic memory, semantic memory, working memory, and attention/processing speed. Baseline tau burden, the average tau intensity across previously defined AD signature regions, and baseline tau spread extent, the proportion of the brain with elevated tau pathology, were quantified for each participant as Tau Index and Tau Spatial Spread, respectively. Tau burden and tau spread were evaluated relative to baseline and longitudinal cognitive performance, as well as longitudinal clinical progression. Tau burden and tau spread extent both significantly correlate with cognitive impairment in symptomatic AD. Tau burden is most strongly correlated with episodic (r = -0.37, p = 0.02) and semantic (r = -0.36, p = 0.02) memory. In contrast, tau spread extent is most strongly correlated with the Knight PACC (r = -0.37, p = 0.01) and attention/processing speed (r = -0.44, p < 0.01), especially in preclinical AD (r = -0.27, p < 0.01). Tau burden captures more variance than tau spread extent in longitudinal change in the Knight PACC, episodic memory, semantic memory, attention/processing speed, and clinical progression. Tau burden strongly relates to baseline episodic and semantic memory, which may reflect that it is heavily weighted by entorhinal tau, a region previously linked to memory processing. In contrast, stronger associations between tau spread extent and baseline attention/processing speed could reflect the inclusion of additional brain regions, particularly the frontal lobe, which support a wider range of cognitive processing. Additionally, tau spread extent is generally more sensitive to baseline preclinical deficits; however, tau burden better estimates future decline across all cognitive domains and clinical symptom onset. Together, these findings suggest complementary utility of evaluating both tau burden and tau spread extent in early AD progression.
Article
Full-text available
People parse continuous experiences at natural breakpoints called event boundaries, which is important for understanding an environment’s causal structure and for responding to uncertainty within it. However, it remains unclear how different forms of uncertainty affect the parsing of continuous experiences and how such uncertainty influences the brain’s processing of ongoing events. We exposed human participants of both sexes (N = 34) to a continuous sequence of semantically meaningless images. We generated sequences from random walks through a graph that grouped images into temporal communities. After learning, we asked participants to segment another sequence at natural breakpoints (event boundaries). Participants segmented the sequence at learned transitions between communities, as well as at novel transitions, suggesting that people can segment temporally extended experiences into events based on learned structure as well as prediction error. Greater segmentation at Novel boundaries was associated with enhanced parietal scalp EEG activity between 250-450 ms after stimulus onset. Multivariate classification of EEG activity showed that Novel and Learned boundaries evoked distinct patterns of neural activity, particularly theta band power in posterior electrodes. Learning also led to distinct neural representations for stimuli within the temporal communities, while neural activity at Learned boundary nodes showed predictive evidence for the adjacent community. The data show that people segment experiences at both learned and novel boundaries and suggest that learned event boundaries trigger retrieval of information about the upcoming community that could underlie anticipation of the next event in a sequence. Significance Statement People make sense of their continuous experience by segmenting it into meaningful units at event boundaries. Event boundaries influence cognitive function in a variety of ways, however it remains unclear how different forms of uncertainty affect the parsing of continuous experiences, and how such uncertainty might influence the brain’s processing of ongoing events. We found that although people segment experiences at both learned and novel boundaries, brain activity diverges rapidly (250-450 ms post-stimulus) in response to different types of event boundaries. The findings suggest the brain can flexibly respond to event boundaries of distinct types, which could support dynamic modulation and updating of neural activity in response to ongoing experience.
Article
Full-text available
A central aspect of episodic memory is the formation of associations between stimuli from different modalities. Current theoretical approaches assume a functional role of ongoing oscillatory power and phase in the theta band (3–7 Hz) for the encoding of crossmodal associations. Furthermore, ongoing activity in the theta range as well as alpha (8–12 Hz) and low beta activity (13–20 Hz) before the presentation of a stimulus is thought to modulate subsequent cognitive processing, including processes that are related to memory. In this study, we tested the hypothesis that pre-stimulus characteristics of low frequency activity are relevant for the successful formation of crossmodal memory. The experimental design that was used specifically allowed for the investigation of associative memory independent from individual item memory. Participants (n = 51) were required to memorize associations between audiovisual stimulus pairs and distinguish them from newly arranged ones consisting of the same single stimuli in the subsequent recognition task. Our results show significant differences in the state of pre-stimulus theta and alpha power between remembered and not remembered crossmodal associations, clearly relating increased power to successful recognition. These differences were positively correlated with memory performance, suggesting functional relevance for behavioral measures of associative memory. Further analysis revealed similar effects in the low beta frequency ranges, indicating the involvement of different pre-stimulus-related cognitive processes. Phase-based connectivity measures in the theta band did not differ between remembered and not remembered stimulus pairs. The findings support the assumed functional relevance of theta band oscillations for the formation of associative memory and demonstrate that an increase of theta as well as alpha band oscillations in the pre-stimulus period is beneficial for the establishment of crossmodal memory.
Article
Full-text available
Two experiments are reported that examined the temporal structure of recall for categorizable word lists by younger and older adults. All participants showed response bursting, in which recall order is clustered by semantic category, with longer interresponse times (IRTs) appearing between categories than within categories. Experiment 1 demonstrated that older adults, even when matched to younger adults in overall accuracy, differed in the rate of increase of between-category IRTs with output position, but not in within-category IRTs. Experiment 2 showed that this interaction is eliminated when the names of the response categories are provided to the participants. Results are interpreted in terms of combined effects of an age-compromised episodic memory system (between-category IRTs) accompanied by a comparatively preserved semantic system (within-category IRTs) in healthy aging.
Article
Full-text available
Presents a spreading-activation theory of human semantic processing, which can be applied to a wide range of recent experimental results. The theory is based on M. R. Quillian's (1967) theory of semantic memory search and semantic preparation, or priming. In conjunction with this, several misconceptions concerning Quillian's theory are discussed. A number of additional assumptions are proposed for his theory to apply it to recent experiments. The present paper shows how the extended theory can account for results of several production experiments by E. F. Loftus, J. F. Juola and R. C. Atkinson's (1971) multiple-category experiment, C. Conrad's (1972) sentence-verification experiments, and several categorization experiments on the effect of semantic relatedness and typicality by K. J. Holyoak and A. L. Glass (1975), L. J. Rips et al (1973), and E. Rosch (1973). The paper also provides a critique of the Rips et al model for categorization judgments. (44 ref)
Article
Full-text available
The idea that synchronous neural activity underlies cognition has driven an extensive body of research in human and animal neuroscience. Yet, insufficient data on intracranial electrical connectivity has precluded a direct test of this hypothesis in a whole-brain setting. Through the lens of memory encoding and retrieval processes, we construct whole-brain connectivity maps of fast gamma (30–100 Hz) and slow theta (3–8 Hz) spectral neural activity, based on data from 294 neurosurgical patients fitted with indwelling electrodes. Here we report that gamma networks desynchronize and theta networks synchronize during encoding and retrieval. Furthermore, for nearly all brain regions we studied, gamma power rises as that region desynchronizes with gamma activity elsewhere in the brain, establishing gamma as a largely asynchronous phenomenon. The abundant phenomenon of theta synchrony is positively correlated with a brain region’s gamma power, suggesting a predominant low-frequency mechanism for inter-regional communication.
Article
Full-text available
Wernicke (1900, as cited in G. H. Eggert, 1977) suggested that semantic knowledge arises from the interaction of perceptual representations of objects and words. The authors present a parallel distributed processing implementation of this theory, in which semantic representations emerge from mechanisms that acquire the mappings between visual representations of objects and their verbal descriptions. To test the theory, they trained the model to associate names, verbal descriptions, and visual representations of objects. When its inputs and outputs are constructed to capture aspects of structure apparent in attribute-norming experiments, the model provides an intuitive account of semantic task performance. The authors then used the model to understand the structure of impaired performance in patients with selective and progressive impairments of conceptual knowledge. Data from 4 well-known semantic tasks revealed consistent patterns that find a ready explanation in the model. The relationship between the model and related theories of semantic representation is discussed.
Article
Full-text available
Previous studies have documented poor recognition memory for faces in patients with semantic dementia. Preserved face recognition memory was found in this study, however, so long as atrophy was confined predominantly to the left temporal lobe. Patients with structural damage to the right temporal lobe were typically impaired, with the status of the hippocampus and parahippocampal gyrus (including the perirhinal cortex) on the right being critical. Two single-case studies of patients with predominantly left temporal lobe pathology confirmed good recognition memory for famous faces, even if semantic knowledge about the celebrities depicted was severely degraded. An effect of semantic knowledge on recognition memory became apparent only when perceptually different photographs of the famous people were used at study and test. These results support the view that new episodic learning typically draws on information from both perceptual and semantic systems.
Article
Intracranial electroencephalography (iEEG), also known as electrocorticography when using subdural grid electrodes or stereotactic EEG when using depth electrodes, is blossoming in various fields of human neuroscience. In this article, we highlight the potentials of iEEG in exploring functions of the human brain while also considering its limitations. The iEEG signal provides anatomically precise information about the selective engagement of neuronal populations at the millimeter scale and the temporal dynamics of their engagement at the millisecond scale. If several nodes of a given network are monitored simultaneously with implanted electrodes, the iEEG signals can also reveal information about functional interactions within and across networks during different stages of neural computation. As such, human iEEG can complement other methods of neuroscience beyond simply replicating what is already known, or can be known, from noninvasive lines of research in humans or from invasive recordings in nonhuman mammalian brains.
Article
Concepts organize our experiences and allow for meaningful inferences in novel situations. Acquiring new concepts requires extracting regularities across multiple learning experiences, a process formalized in mathematical models of learning. These models posit a computational framework that has increasingly aligned with the expanding repertoire of functions associated with the hippocampus. Here, we propose the Episodes-to-Concepts (EpCon) theoretical model of hippocampal function in concept learning and review evidence for the hippocampal computations that support concept formation including memory integration, attentional biasing, and memory-based prediction error. We focus on recent studies that have directly assessed the hippocampal role in concept learning with an innovative approach that combines computational modeling and sophisticated neuroimaging measures. Collectively, this work suggests that the hippocampus does much more than encode individual episodes; rather, it adaptively transforms initially-encoded episodic memories into organized conceptual knowledge that drives novel behavior.
Article
The dilemma that neurotheorists face is that (1) detailed biophysical models that can be constrained by direct measurements, while being of great importance, offer no immediate insights into cognitive processes in the brain, and (2) high-level abstract cognitive models, on the other hand, while relevant for understanding behavior, are largely detached from neuronal processes and typically have many free, experimentally unconstrained parameters that have to be tuned to a particular data set and, hence, cannot be readily generalized to other experimental paradigms. In this contribution, we propose a set of “first principles” for neurally inspired cognitive modeling of memory retrieval that has no biologically unconstrained parameters and can be analyzed mathematically both at neuronal and cognitive levels. We apply this framework to the classical cognitive paradigm of free recall. We show that the resulting model accounts well for puzzling behavioral data on human participants and makes predictions that could potentially be tested with neurophysiological recording techniques.
Article
People often forget information because they fail to effectively encode it. Here, we test the hypothesis that targeted electrical stimulation can modulate neural encoding states and subsequent memory outcomes. Using recordings from neurosurgical epilepsy patients with intracranially implanted electrodes, we trained multivariate classifiers to discriminate spectral activity during learning that predicted remembering from forgetting, then decoded neural activity in later sessions in which we applied stimulation during learning. Stimulation increased encoding-state estimates and recall if delivered when the classifier indicated low encoding efficiency but had the reverse effect if stimulation was delivered when the classifier indicated high encoding efficiency. Higher encoding-state estimates from stimulation were associated with greater evidence of neural activity linked to contextual memory encoding. In identifying the conditions under which stimulation modulates memory, the data suggest strategies for therapeutically treating memory dysfunction.
Article
Neural networks that span the medial temporal lobe (MTL), prefrontal cortex, and posterior cortical regions are essential to episodic memory function in humans. Encoding and retrieval are supported by the engagement of both distinct neural pathways across the cortex and common structures within the medial temporal lobes. However, the degree to which memory performance can be determined by neural processing that is common to encoding and retrieval remains to be determined. To identify neural signatures of successful memory function, we administered a delayed free-recall task to 187 neurosurgical patients implanted with subdural or intraparenchymal depth electrodes. We developed multivariate classifiers to identify patterns of spectral power across the brain that independently predicted successful episodic encoding and retrieval. During encoding and retrieval, patterns of increased high frequency activity in prefrontal, MTL, and inferior parietal cortices, accompanied by widespread decreases in low frequency power across the brain predicted successful memory function. Using a cross-decoding approach, we demonstrate the ability to predict memory function across distinct phases of the free-recall task. Furthermore, we demonstrate that classifiers that combine information from both encoding and retrieval states can outperform task-independent models. These findings suggest that the engagement of a core memory network during either encoding or retrieval shapes the ability to remember the past, despite distinct neural interactions that facilitate encoding and retrieval.