Neural mechanisms of semantic memory

Multimodal study of multilevel pulvino-temporal connections: a new piece in the puzzle of lexical retrieval networks

Advanced methods of imaging and mapping the healthy and lesioned brain have allowed for the identification of the cortical nodes and white matter tracts supporting the dual neurofunctional organization of language networks in a dorsal phonological and a ventral semantic stream. Much less understood are the anatomical correlates of the interaction between the two streams; one hypothesis being that of a subcortically mediated interaction, through crossed cortico-striato-thalamo-cortical and cortico-thalamo-cortical loops. In this regard, the pulvinar is the thalamic subdivision that has most regularly appeared as implicated in the processing of lexical retrieval. However, descriptions of its connections with temporal (language) areas remain scarce. Here we assess this pulvino-temporal connectivity using a combination of state-of-the-art techniques: white matter stimulation in awake surgery and postoperative diffusion MRI (n = 4), virtual dissection from the Human Connectome Project 3 and 7 T datasets (n = 172) and operative microscope-assisted post-mortem fibre dissection (n = 12). We demonstrate the presence of four fundamental fibre contingents: (i) the anterior component (Arnold's bundle proper) initially described by Arnold in the 19th century and destined to the anterior temporal lobe; (ii) the optic radiations-like component, which leaves the pulvinar accompanying the optical radiations and reaches the posterior basal temporal cortices; (iii) the lateral component, which crosses the temporal stem orthogonally and reaches the middle temporal gyrus; and (iv) the auditory radiations-like component, which leaves the pulvinar accompanying the auditory radiations to the superomedial aspect of the temporal operculum, just posteriorly to Heschl's gyrus. Each of those components might correspond to a different level of information processing involved in the lexical retrieval process of picture naming.

Neural manifestation of L2 novel concept acquisition from multi-contexts via both episodic memory and semantic memory systems

This study aims to examine the process of L2 novel word learning through the combination of episodic and semantic memory, and how the process differs between the formation of thematic and taxonomic relations. The major approach adopted was observing the neural effects of word learning, which is manifested in the N400 from event-related potentials (ERPs). Eighty-eight participants were recruited for the experiment. In the learning session, L2 contextual discourses related to novel words were learned by participants. In the testing session, discourses embedded with incongruous and congruous novel words in the final position were used for participants to judge the congruency which affected the N400 neural activity. The results showed that both recurrent and new-theme discourses elicited significant N400 effects, while taxonomic sentences did not. These results confirmed the formation of episodic and semantic memory during L2 new word learning, in which semantic memory was mainly supported by thematic relations.

Semantic memory retrieval circuit: role of pre-SMA, caudate, and thalamus

We propose that pre-supplementary motor area (pre-SMA)-thalamic interactions govern processes fundamental to semantic retrieval of an integrated object memory. At the onset of semantic retrieval, pre-SMA initiates electrical interactions between multiple cortical regions associated with semantic memory subsystems encodings as indexed by an increase in theta-band EEG power. This starts between 100-150 ms after stimulus presentation and is sustained throughout the task. We posit that this activity represents initiation of the object memory search, which continues in searching for an object memory. When the correct memory is retrieved, there is a high beta-band EEG power increase, which reflects communication between pre-SMA and thalamus, designates the end of the search process and resultant in object retrieval from multiple semantic memory subsystems. This high beta signal is also detected in cortical regions. This circuit is modulated by the caudate nuclei to facilitate correct and suppress incorrect target memories.

Altered brain function underlying verbal memory encoding and retrieval in psychotic major depression

Psychotic major depression (PMD) is associated with deficits in verbal memory as well as other cognitive impairments. This study investigated brain function in individuals with PMD during a verbal declarative memory task. Participants included 16 subjects with PMD, 15 subjects with non-psychotic major depression (NPMD) and 16 healthy controls (HC). Functional magnetic resonance imaging (fMRI) data were acquired while subjects performed verbal memory encoding and retrieval tasks. During the explicit encoding task, subjects semantically categorized words as either "man-made" or "not man-made." For the retrieval task, subjects identified whether words had been presented during the encoding task. Functional MRI data were processed using SPM5 and a group by condition ANOVA. Clusters of activation showing either a significant main effect of group or an interaction of group by condition were further examined using t-tests to identify group differences. During the encoding task, the PMD group showed lower hippocampus, insula, and prefrontal activation compared to HC. During the retrieval task, the PMD group showed lower recognition accuracy and higher prefrontal and parietal cortex activation compared to both HC and NPMD groups. Verbal retrieval deficits in PMD may be associated with deficient hippocampus function during encoding. Increased brain activation during retrieval may reflect an attempt to compensate for encoding deficits.

The role of the pulvinar in resolving competition between memory and visual selection: a functional connectivity study

Memory and attention interact. Information held in working memory (WM) can bias visual selection toward matching stimuli in a subsequent search display, while a search target that is different from the memory stimulus can interfere with its subsequent recognition. In recent fMRI studies, the pulvinar has been consistently shown to have an enhanced response when an item in WM matches a search target and a reduced response when the WM item matches a distracter in search. Here we used Granger causality analysis to help understand the role of the pulvinar in resolving competition between memory and selection processes. Across three experiments the results showed increased coupling between the pulvinar and the ipsilateral superior frontal gyrus, contralateral temporal-parietal junction (TPJ) and calcarine sulcus when a visual search distracter matched the item held in memory. This connection pattern suggests that the pulvinar suppresses visual responses to the target when a contralateral distracter contains information held in working memory. We propose that this suppression acts to protect the memory item from interference arising from information associated with the search target. Consistent with this proposal we showed that the strength of the thalamus-to-visual connection predicted performance on a subsequent memory test. The data therefore suggest that the thalamus modulates bottom up processing in sensory cortex to minimize interference to WM content.

An integrated neural model of semantic memory, lexical retrieval and category formation, based on a distributed feature representation

This work presents a connectionist model of the semantic-lexical system. Model assumes that the lexical and semantic aspects of language are memorized in two distinct stores, and are then linked together on the basis of previous experience, using physiological learning mechanisms. Particular characteristics of the model are: (1) the semantic aspects of an object are described by a collection of features, whose number may vary between objects. (2) Individual features are topologically organized to implement a similarity principle. (3) Gamma-band synchronization is used to segment different objects simultaneously. (4) The model is able to simulate the formation of categories, assuming that objects belong to the same category if they share some features. (5) Homosynaptic potentiation and homosynaptic depression are used within the semantic network, to create an asymmetric pattern of synapses; this allows a different role to be assigned to shared and distinctive features during object reconstruction. (6) Features which frequently occurred together, and the corresponding word-forms, become linked via reciprocal excitatory synapses. (7) Features in the semantic network tend to inhibit words not associated with them during the previous learning phase. Simulations show that, after learning, presentation of a cue can evoke the overall object and the corresponding word in the lexical area. Word presentation, in turn, activates the corresponding features in the sensory-motor areas, recreating the same conditions occurred during learning, according to a grounded cognition viewpoint. Several words and their conceptual description can coexist in the lexical-semantic system exploiting gamma-band time division. Schematic exempla are shown, to illustrate the possibility to distinguish between words representing a category, and words representing individual members and to evaluate the role of gamma-band synchronization in priming. Finally, the model is used to simulate patients with focalized lesions, assuming a damage of synaptic strength in specific feature areas. Results are critically discussed in view of future model extensions and application to real objects. The model represents an original effort to incorporate many basic ideas, found in recent conceptual theories, within a single quantitative scaffold.

A computational model of the lexical-semantic system based on a grounded cognition approach

This work presents a connectionist model of the semantic-lexical system based on grounded cognition. The model assumes that the lexical and semantic aspects of language are memorized in two distinct stores. The semantic properties of objects are represented as a collection of features, whose number may vary among objects. Features are described as activation of neural oscillators in different sensory-motor areas (one area for each feature) topographically organized to implement a similarity principle. Lexical items are represented as activation of neural groups in a different layer. Lexical and semantic aspects are then linked together on the basis of previous experience, using physiological learning mechanisms. After training, features which frequently occurred together, and the corresponding word-forms, become linked via reciprocal excitatory synapses. The model also includes some inhibitory synapses: features in the semantic network tend to inhibit words not associated with them during the previous learning phase. Simulations show that after learning, presentation of a cue can evoke the overall object and the corresponding word in the lexical area. Moreover, different objects and the corresponding words can be simultaneously retrieved and segmented via a time division in the gamma-band. Word presentation, in turn, activates the corresponding features in the sensory-motor areas, recreating the same conditions occurring during learning. The model simulates the formation of categories, assuming that objects belong to the same category if they share some features. Simple exempla are shown to illustrate how words representing a category can be distinguished from words representing individual members. Finally, the model can be used to simulate patients with focalized lesions, assuming an impairment of synaptic strength in specific feature areas.

Smoking is associated with reduced cortical regional gray matter density in brain regions associated with incipient Alzheimer disease

OBJECTIVES

The results of observational studies suggest that smoking increases the risk of Alzheimer disease (AD). The authors designed this study to determine if older people who smoke have decreased gray matter density in brain regions associated with incipient AD.

METHODS

The authors recruited 39 pairs (N = 78) of smokers/never-smokers 70 to 83 years of age who were matched for age, sex, education, and handedness. Participants were free of clinically significant cognitive impairment, depression, stroke, or other serious medical conditions. Gray matter density was determined by voxel-based morphometry using statistical parametric mapping of T1-weighted magnetic resonance images.

RESULTS

Smokers had decreased gray matter density in the posterior cingulum and precuneus (bilateral), right thalamus, and frontal cortex (bilateral) compared with never-smokers.

CONCLUSIONS

Smoking is associated with decreased gray matter density in brain regions previously associated with incipient AD. Longitudinal investigations are required to clarify whether these changes are progressive in nature.

Wechsler Memory Scale Revised Edition: neural correlates of the visual paired associates subtest adapted for fMRI

Memory deficits in neurological and psychiatric patients are evaluated by neuropsychological tests such as the Wechsler Memory Scale Revised Edition (WMS-R). Neuropsychological data from patients with circumscribed lesions point to single elements of the underlying neural network but fail to identify its whole extent. We report the fMRI adaptation of a subtest of the WMS-R, the Visual Paired Associates. Fifteen healthy, right-handed male volunteers were studied using a 1.5T MRI scanner. The encoding of the combination between a shape and a colour, the assessment of the retrieval of this combination immediately after encoding took place, and the underlying network employed during retrieval a second time after approximately 25 min were investigated. The results show a fronto-parieto-occipital network with left frontal accentuation for encoding and a fronto-parieto-occipital network for immediate and delayed retrieval. Noteworthy is the specific role of the thalamus. During immediate retrieval, the thalamus showed significant bilateral activation; during delayed retrieval, there was no significant activation. The thalami are part of an extended hippocampal-diencephalic system which is critical for efficient encoding and normal retrieval of new episodic information. We describe the probability of thalamocortical connections during retrieval based on the Thalamus Connectivity Atlas. The cerebellum showed significant activation in all conditions; its part in higher cognitive functions such as memory was thereby confirmed.

Neural substrates of semantic memory

Semantic memory is described as the storage of knowledge, concepts, and information that is common and relatively consistent across individuals (e.g., memory of what is a cup). These memories are stored in multiple sensorimotor modalities and cognitive systems throughout the brain (e.g., how a cup is held and manipulated, the texture of a cup's surface, its shape, its function, that is related to beverages such as coffee, and so on). Our ability to engage in purposeful interactions with our environment is dependent on the ability to understand the meaning and significance of the objects and actions around us that are stored in semantic memory. Theories of the neural basis of the semantic memory of objects have produced sophisticated models that have incorporated to varying degrees the results of cognitive and neural investigations. The models are grouped into those that are (1) cognitive models, where the neural data are used to reveal dissociations in semantic memory after a brain lesion occurs; (2) models that incorporate both cognitive and neuroanatomical information; and (3) models that use cognitive, neuroanatomic, and neurophysiological data. This review highlights the advances and issues that have emerged from these models and points to future directions that provide opportunities to extend these models. The models of object memory generally describe how category and/or feature representations encode for object memory, and the semantic operations engaged in object processing. The incorporation of data derived from multiple modalities of investigation can lead to detailed neural specifications of semantic memory organization. The addition of neurophysiological data can potentially provide further elaboration of models to include semantic neural mechanisms. Future directions should incorporate available and newly developed techniques to better inform the neural underpinning of semantic memory models.

Effect of Antipsychotic Medication Use on Memory in Patients With Alzheimer’s Disease: Assessing the Potential Risk for Accelerated Recent Autobiographical Memory Loss

Autobiographical memory loss is a common and disturbing problem for individuals with Alzheimer's disease (AD). Patients with AD who are taking antipsychotic medications may be at further risk for loss of recent autobiographical memory because of the potential anticholinergic side effects of antipsychotics. The purpose of this post hoc, descriptive study was to compare the recent autobiographical memory scores of patients with AD taking antipsychotics to those who were not taking antipsychotics. The study population was composed of 35 patients with moderate-stage AD. Patients who were taking antipsychotics scored significantly worse on a recent autobiographical memory measure compared with patients who were not taking antipsychotics. This study provides further evidence for judicious use of antipsychotic medications with AD patients.

The Semantic Object Retrieval Test (SORT) in Amnestic Mild Cognitive Impairment

BACKGROUND

Between 10% and 15% of patients with the amnestic variety of Mild Cognitive Impairment (MCI) convert to Alzheimer disease (AD) per year.

OBJECTIVE

Characterize cognitive markers that may herald conversion from MCI to AD and directly assess semantic memory in patients meeting criteria for amnestic MCI.

DESIGN

Thirty-five amnestic MCI patients and 121 healthy aging controls enrolled at an Alzheimer Disease Center received a battery of standard neuropsychologic tests, and the Semantic Object Retrieval Test (SORT), a test that we have developed for the assessment of semantic memory and subsequent name production, and that has been shown to be able to differentiate between normals and patients with AD.

RESULTS

On the basis of normative data from the SORT, the MCI subjects could be divided into 2 groups: 10 patients (29%) with a significant semantic impairment (SI+) and 25 without a semantic memory deficit (SI-). There was a significant correlation between all SORT variables and performance on the Boston Naming Test. In this MCI population, significantly impaired SORT performance was associated with a relative decrease in performance on tests of frontal lobe functions, although disruption of thalamic-related processes cannot be excluded as an etiology for semantic memory impairment.

CONCLUSIONS

The SORT is a specific test of semantic memory, and is a sensitive measure of semantic memory deficits in patients who otherwise meet criteria for amnestic MCI. Using this specific assessment tool, a significant number of MCI patients were found to have semantic memory deficits. As these patients may be early in the course of possible progression toward dementia, the SORT or other tests of semantic memory may provide important diagnostic or prognostic information in patients with MCI.

Neural correlates of verbal semantic memory in patients with temporal lobe epilepsy

Functional imaging data suggest that the core network engaged in verbal semantic memory (SM) processing encompasses frontal and temporal lobe structures, with a strong left lateralization in normal right handers. The impact of long term temporal lobe epilepsy (TLE) on this network has only partly been elucidated. We studied verbal SM in 50 patients with chronic, intractable TLE (left TLE=26, right TLE=24) and 35 right handed normal controls using a verbal fMRI semantic decision paradigm. All patients had language lateralized to the left hemisphere, as verified by the intracarotid amobarbital procedure. Within and between group analyses showed remarkable, group-specific activation profiles. The control group activated frontal and temporal areas bilaterally, with a strong left predominance. Left TLE patients showed a shift of activations of left frontal and medial temporal areas to homologous regions in the right hemisphere. Furthermore, left TLE subjects utilized subcortical structures such as the thalamus and putamen to accomplish the verbal SM task. Contrastively, the activation pattern of right TLE patients resembled that of normal controls, but exhibited "hypofrontality" with a shift from frontal to posterior regions in the temporal, parietal and occipital lobe. Our results show that chronic epileptic activity originating from temporal seizure foci is associated with an alteration of neural circuits which support semantic language processing and that side of seizure focus has a specific impact on the resulting activation network. These findings presumably result from morphological changes and from functional reorganization which are both inherent to chronic TLE.

The effects of masking on the activation of auditory-associated cortex during speech listening in white noise

CONCLUSIONS

Noise-induced masking has different effects on the two hemispheres during speech listening. Auditory-associated cortices in the left hemisphere were more affected by masking than the right side. However, activation of primary and secondary auditory cortices was not affected in both sides under the masking with high signal to noise ratio.

OBJECTIVES

The purpose of this study was to investigate the effects of masking on the central auditory system during speech listening in white noise.

MATERIALS AND METHODS

Twelve healthy young subjects with normal hearing participated in this study. Functional magnetic resonance imaging (fMRI) was performed while subjects were listening to speech sounds alone and speech plus white noise binaurally.

RESULTS

In humans, the activation of several regions including the middle parts of the superior and middle temporal gyri, parahippocampal gyrus, cuneus and thalamus of the left hemisphere was significantly reduced under the masking paradigm with +5 dB signal to noise ratio. In addition, reduced activation was also found at the lingual gyrus, anterior and middle parts of the superior temporal gyrus (STG), uncus, fusiform gyrus, and inferior frontal gyrus of the right hemisphere during masking.