Impaired inference in a case of developmental amnesia

The transitive inference task to study the neuronal correlates of memory-driven decision making: A monkey neurophysiology perspective

A vast amount of literature agrees that rank-ordered information as A>B>C>D>E>F is mentally represented in spatially organized schemas after learning. This organization significantly influences the process of decision-making, using the acquired premises, i.e. deciding if B is higher than D is equivalent to comparing their position in this space. The implementation of non-verbal versions of the transitive inference task has provided the basis for ascertaining that different animal species explore a mental space when deciding among hierarchically organized memories. In the present work, we reviewed several studies of transitive inference that highlighted this ability in animals and, consequently, the animal models developed to study the underlying cognitive processes and the main neural structures supporting this ability. Further, we present the literature investigating which are the underlying neuronal mechanisms. Then we discuss how non-human primates represent an excellent model for future studies, providing ideal resources for better understanding the neuronal correlates of decision-making through transitive inference tasks.

Exploring the sociodemographic, clinical and neuropsychological factors associated with relational memory in schizophrenia

INTRODUCTION

The Transverse Patterning (TP) task has been used to measure episodic relational memory (RM) deficits in clinical populations. Individuals with schizophrenia often fail to learn TP with standard, and sometimes extensive training. Identifying the differences between TP learners and non-learners can improve our understanding of successful TP performance and its underlying mechanisms, which may help improve interventions aimed at ameliorating RM performance. We investigated sociodemographic, clinical and neuropsychological factors associated with TP performance in schizophrenia.

METHODS

Sixty-six participants with schizophrenia completed a semantically rich and a relational-binding dependent version of the TP task and reported on their task awareness and strategy use.

RESULTS

Twenty-six participants failed to learn the task rules after extensive training. Learners had superior verbal, visual and working memory, executive functions and overall cognitive functioning compared to non-learners. Learners also had superior awareness of task rules and pairs relationships and used elaborated cognitive strategies more often.

CONCLUSIONS

Our results support previous findings that some individuals with schizophrenia show RM impairment even with extensive TP training. We shed light on neuropsychological and metacognitive factors associated with TP performance. This knowledge could enhance interventions targeted to improve relational memory in schizophrenia when extensive training fails.

Inverse forgetting in unconscious episodic memory

Forming memories of experienced episodes calls upon the episodic memory system. Episodic encoding may proceed with and without awareness of episodes. While up to 60% of consciously encoded episodes are forgotten after 10 h, the fate of unconsciously encoded episodes is unknown. Here we track over 10 h, which are filled with sleep or daytime activities, the retention of unconsciously and consciously experienced episodes. The episodes were displayed in cartoon clips that were presented weakly and strongly masked for conscious and unconscious encoding, respectively. Clip retention was tested for distinct clips directly after encoding, 3 min and 10 h after encoding using a forced-choice test that demands deliberate responses in both consciousness conditions. When encoding was conscious, retrieval accuracy decreased by 25% from 3 min to 10 h, irrespective of sleep or wakefulness. When encoding was unconscious, retrieval accuracy increased from 3 min to 10 h and depended on sleep. Hence, opposite to the classic forgetting curve, unconsciously acquired episodic memories strengthen over time and hinge on sleep on the day of learning to gain influence over human behavior.

The Role of the Ventromedial Prefrontal Cortex and Basal Forebrain in Relational Memory and Inference

The ventromedial prefrontal cortex (vmPFC) is involved in diverse cognitive operations, from inhibitory control to processing of semantic schemas. When accompanied by damage to the basal forebrain, vmPFC lesions can also impair relational memory, the ability to form and recall relations among items. Impairments in establishing direct relations among items (e.g., A is related to B, B is related to C) can also hinder the transitive processing of indirect relationships (e.g., inferring that A and C are related through direct relations that each contain B). Past work has found that transitive inference improves when the direct relations are organized within an existing knowledge structure, or schema. This type of semantic support is most effective for individuals whose relational memory deficits are mild (e.g., healthy age-related decline) rather than pronounced (e.g., hippocampal amnesia, amnestic mild cognitive impairment). Given that vmPFC damage can produce both relational memory and schema processing deficits, such damage may pose a particular challenge in establishing the type of relational structure required for transitive inference, even when supported by preexisting knowledge. To examine this idea, we tested individuals with lesions to the mPFC on multiple conditions that varied in pre-experimental semantic support and explored the extent to which they could identify both previously studied (direct) and novel transitive (indirect) relations. Most of the mPFC cases showed marked transitive inference deficits and even showed impaired knowledge of preexisting, direct, semantic relations, consistent with disruptions to schema-related processes. However, one case with more dorsal mPFC damage showed preserved ability to identify direct relations and make novel inferences, particularly when pre-experimental knowledge could be used to support performance. These results suggest that damage to the mPFC and basal forebrain can impede establishment of ad hoc relational schemas upon which transitive inference is based, but that appealing to prior knowledge may still be useful for those neurological cases that have some degree of preserved relational memory.

Larger capacity for unconscious versus conscious episodic memory

Episodic memory is the memory for experienced events. A peak competence of episodic memory is the mental combination of events to infer commonalities. Inferring commonalities may proceed with and without consciousness of events. Yet what distinguishes conscious from unconscious inference? This question inspired nine experiments that featured strongly and weakly masked cartoon clips presented for unconscious and conscious inference. Each clip featured a scene with a visually impenetrable hiding place. Five animals crossed the scene one-by-one consecutively. One animal trajectory represented one event. The animals moved through the hiding place, where they might linger or not. The participants' task was to observe the animals' entrances and exits to maintain a mental record of which animals hid simultaneously. We manipulated information load to explore capacity limits. Memory of inferences was tested immediately, 3.5 or 6 min following encoding. The participants retrieved inferences well when encoding was conscious. When encoding was unconscious, the participants needed to respond intuitively. Only habitually intuitive decision makers exhibited a significant delayed retrieval of inferences drawn unconsciously. Their unconscious retrieval performance did not drop significantly with increasing information load, while conscious retrieval performance dropped significantly. A working memory network, including hippocampus, was activated during both conscious and unconscious inference and correlated with retrieval success. An episodic retrieval network, including hippocampus, was activated during both conscious and unconscious retrieval of inferences and correlated with retrieval success. Only conscious encoding/retrieval recruited additional brain regions outside these networks. Hence, levels of consciousness influenced the memories' behavioral impact, memory capacity, and the neural representational code.

Representation of viewed and recalled film clips in patterns of brain activity in a person with developmental amnesia

As clear memories transport us back into the past, the brain also revives prior patterns of neural activity, a phenomenon known as neural reactivation. While growing evidence indicates a link between neural reactivation and typical variations in memory performance in healthy individuals, it is unclear how and to what extent reactivation is disrupted by a memory disorder. The current study characterizes neural reactivation in a case of amnesia using Multivoxel Pattern Analysis (MVPA). We tested NC, an individual with developmental amnesia linked to a diencephalic stroke, and 19 young adult controls on a functional magnetic resonance imaging (fMRI) task during which participants viewed and recalled short videos multiple times. An encoding classifier trained and tested to identify videos based on brain activity patterns elicited at perception revealed superior classification in NC. The enhanced consistency in stimulus representation we observed in NC at encoding was accompanied by an absence of multivariate repetition suppression, which occurred over repeated viewing in the controls. Another recall classifier trained and tested to identify videos during mental replay indicated normal levels of classification in NC, despite his poor memory for stimulus content. However, a cross-condition classifier trained on perception trials and tested on mental replay trials-a strict test of reactivation-revealed significantly poorer classification in NC. Thus, while NC's brain activity was consistent and stimulus-specific during mental replay, this specificity did not reflect the reactivation of patterns elicited at perception to the same extent as controls. Fittingly, we identified brain regions for which activity supported stimulus representation during mental replay to a greater extent in NC than in controls. This activity was not modeled on perception, suggesting that compensatory patterns of representation based on generic knowledge can support consistent mental constructs when memory is faulty. Our results reveal several ways in which amnesia impacts distributed patterns of stimulus representation during encoding and retrieval.

Gradual learning and inflexible strategy use in amnesia: Evidence from case H.C

The value of case studies in informing our understanding of dissociations and interactions in memory was recognized early on by Endel Tulving, whose comprehensive work with the amnesic case K.C. helped to confirm distinctions between episodic and semantic memory. Following in this tradition, we examined memory and the use of cognitive strategies in the developmental amnesic case H.C., a young woman with structural abnormalities in the extended hippocampal system (Rosenbaum et al., 2014). H.C. was tested on two tasks, transitivity and transverse patterning, that each required learning the relations among items, and for the former, also examined the ability to make inferences across sets of relations. H.C. was tested across multiple sessions and demonstrated two seemingly contradictory patterns of performance: evidence of gradual learning, yet an inability to flexibly switch to a cognitive strategy that may otherwise benefit performance. Specifically, on the transitivity task, H.C. showed gradual learning of novel relations that led to successful inferential performance. On transverse patterning, H.C. showed some gradual learning of the relations among the objects across sessions, and expressed knowledge that the task followed 'rock-paper-scissors' rules. However, H.C. did not benefit from a unitization strategy, which had shown previous success with other amnesic cases (D'Angelo et al., 2015; Ryan, Moses, Barense, & Rosenbaum, 2013). H.C.'s over-reliance on 'rock-paper-scissors' rules, even in the face of alternate strategies, is suggestive of an inability to enact cognitive flexibility. Poor performance thus may have resulted from interference from the experimentally presented strategy on her self-imposed strategy. The present findings echo work reported by Tulving in case K.C. (Tulving, Hayman, & Macdonald, 1991). Whereas neurologically intact individuals may rely on the functions of the hippocampal system to rapidly learn new information and resolve interference, some individuals with hippocampal amnesia may learn information gradually, but such learning is particularly prone to interference, resulting in an inability to flexibly adapt to changes in the learning conditions in order to optimize performance.

The effect of repetition on pronoun resolution in patients with memory impairment

Referring to things in the world - that woman, her idea, she - is a central component of language. Understanding reference requires the listener to keep track of the unfolding discourse history while integrating multiple sources of information to interpret the speech stream as it unfolds in time. Pronouns are a common way to establish reference. But due to their impoverished form, to understand them listeners must relate features of the pronoun (e.g., gender, animacy) with existing representations of potential discourse referents. Successful referential processing seems to place demands on memory. In a previous study, patients with hippocampal amnesia and healthy participants listened to short stories as their eye movements were monitored. When interpreting ambiguous pronouns, healthy participants demonstrated order-of-mention effects, whereby ambiguous pronouns are interpreted as referring to the first-mentioned referent in the story. By contrast, memory-impaired patients exhibited significant disruptions in their ability to use information about which character had been mentioned first to interpret pronouns. Repetition of the most salient information is a common clinical recommendation for improving pronoun resolution and communication in individuals with memory disorders (e.g., Alzheimer's disease) but this recommendation lacks an evidentiary basis. The present study seeks to determine whether the pronoun resolution performance of hippocampal patients can be improved, by repetition of the target referent, increasing its salience. Results indicate that patients with hippocampal damage demonstrate improved processing of pronouns following repetition of the target referent, but benefit from this repetition to a significantly smaller degree compared to healthy participants. These results provide further evidence for the role of the hippocampal-dependent memory system in language processing and point to the need for empirically tested communication interventions.

Coherence and congruency mediate medial temporal and medial prefrontal activity during event construction

The precise roles of the hippocampus (HPC) and medial prefrontal cortex (mPFC) in initially constructing imagined events remains unclear. HPC activity during imagination may be modulated by mnemonic load, given its role in working memory for complex materials, and/or by the semantic relatedness (i.e. congruency) between items and their context. MPFC activation may track with congruency or mnemonic load, given the role of ventral mPFC in schema processing and the dorsal mPFC in working memory for social information. Sixteen healthy adults (M age = 22.3) underwent an event construction task, wherein participants were provided with a context and item words and imagined an event, forming as many inter-item associations as possible among the items. The stimuli varied by set size and by normatively-defined congruence (normative congruency) to explore their effects on HPC and mPFC activity and functional connectivity. We observed HPC connectivity during event construction in general, whereas dorsal mPFC connectivity occurred during imagining only at higher set sizes. Moreover, anterior hippocampal activity correlated positively with increasing coherence between items during imagining, suggesting that the anterior HPC is sensitive to the relational demands of constructing a novel event. Parahippocampal, hippocampal, temporal pole, and mPFC activity tracked only with individual differences in subjective ratings of congruency of imagined events, which may contribute to construction by retrieving existing schema-related information. Collectively, these findings provide new insights into the factors that modulate HPC and mPFC activity when constructing mental simulations.

The hippocampus is critical for value-based decisions guided by dissociative inference

The hippocampus supports flexible decision-making through memory integration: bridging across episodes and inferring associations between stimuli that were never presented together ('associative inference'). A pre-requisite for memory integration is flexible representations of the relationships between stimuli within episodes (AB) but also of the constituent units (A,B). Here we investigated whether the hippocampus is required for parsing experienced episodes into their constituents to infer their re-combined within-episode associations ('dissociative inference'). In three experiments male rats were trained on an appetitive conditioning task using compound auditory stimuli (AB+, BA+, CD-, DC-). At test either the compound or individual stimuli were presented as well as new stimuli. Rats with hippocampal lesions acquired and retained the compound discriminations as well as controls. Single constituent stimuli (A, B, C, D) were presented for the first time at test, so the only value with which they could be associated was the one from the compound to which they belonged. Controls inferred constituent tones' corresponding values while hippocampal rats did not, treating them as merely familiar stimuli with no associated value. This finding held whether compound training occurred before or after hippocampal lesions, suggesting that hippocampus-dependent inferential processes more likely occur at retrieval. The findings extend recent discoveries about the role of the hippocampus in intrinsic value representation, demonstrating hippocampal contributions to allocating value from primary rewards to individual stimuli. Importantly, we discovered that dissociative inferences serve to restructure or reparse patterns of directly acquired associations when animals are faced with environmental changes and need to extract relevant information from a multiplex memory. The hippocampus is critical for this fundamental flexible use of associations.

Superior explicit memory despite severe developmental amnesia: In-depth case study and neural correlates

The acquisition of new semantic memories is sometimes preserved in patients with hippocampal amnesia. Robust evidence for this comes from case reports of developmental amnesia suggesting that low-to-normal levels of semantic knowledge can be achieved despite compromised episodic learning. However, it is unclear whether this relative preservation of semantic memory results from normal acquisition and retrieval or from residual episodic memory, combined with effortful repetition. Furthermore, lesion studies have mainly focused on the hippocampus itself, and have seldom reported the state of structures in the extended hippocampal system. Preserved components of this system may therefore mediate residual episodic abilities, contributing to the apparent semantic preservation. We report an in-depth study of Patient KA, a 27-year-old man who had severe hypoxia at birth, in which we carefully explored his residual episodic learning abilities. We used novel speeded recognition paradigms to assess whether KA could explicitly acquire and retrieve new context-free memories. Despite a pattern of very severe amnesia, with a 44-point discrepancy between his intelligence and memory quotients, KA exhibited normal-to-superior levels of knowledge, even under strict time constraints. He also exhibited normal-to-superior recognition memory for new material, again under strict time constraints. Multimodal neuroimaging revealed an unusual pattern of selective atrophy within each component of the extended hippocampal system, contrasting with the preservation of anterior subhippocampal cortices. A cortical thickness analysis yielded a pattern of thinner but also thicker regional cortices, pointing toward specific temporal lobe reorganization following early injury. We thus report the first case of superior explicit learning and memory in a severe case of amnesia, raising important questions about how such knowledge can be acquired.

Dissociable contributions of thalamic nuclei to recognition memory: novel evidence from a case of medial dorsal thalamic damage

The thalamic nuclei are thought to play a critical role in recognition memory. Specifically, the anterior thalamic nuclei and medial dorsal nuclei may serve as critical output structures in distinct hippocampal and perirhinal cortex systems, respectively. Existing evidence indicates that damage to the anterior thalamic nuclei leads to impairments in hippocampal-dependent tasks. However, evidence for the opposite pattern following medial dorsal nuclei damage has not yet been identified. In the present study, we investigated recognition memory in NC, a patient with relatively selective medial dorsal nuclei damage, using two object recognition tests with similar foils: a yes/no (YN) test that requires the hippocampus, and a forced choice corresponding test (FCC) that is supported by perirhinal cortex. NC performed normally in the YN test, but was impaired in the FCC test. Critically, FCC performance was impaired only when the study-test delay period was filled with interference. We interpret these results in the context of the representational–hierarchical model, which predicts that memory deficits following damage to the perirhinal system arise due to increased vulnerability to interference. These data provide the first evidence for selective deficits in a task that relies on perirhinal output following damage to the medial dorsal nuclei, providing critical evidence for dissociable thalamic contributions to recognition memory.