Assessing the working memory network: studies with functional magnetic resonance imaging and structural equation modeling

A considerable body of evidence supports the notion that the neurofunctional substrate of working memory is not only related to the integrity of the prefrontal cortex, but also to the concerted interplay of widespread interacting networks including the parietal cortex, subcortical regions and cerebellar areas. Modern functional brain imaging techniques such as functional magnetic resonance imaging (fMRI) have provided a detailed picture of functional neuroanatomy subserving working memory functions. Most of the earlier functional studies were directed toward the identification of brain areas subserving specific cognitive domains in terms of a functional segregation. More recently, different multivariate techniques were employed to specifically address measures of functional or effective connectivity. Structural equation modeling (SEM) or path analysis is one of the most often used methods to model interactions among covarying brain areas in an explicitly model-based approach. The present review will focus on basic methodological issues of SEM for the analysis of fMRI datasets in studies of working memory. Aside from a discussion of previous studies and their essential findings, advanced methodological issues and caveats as well as future perspectives of the method will be addressed.     

Ventrolateral prefrontal cortex activity associated with individual differences in arbitrary delayed paired-association learning performance: A functional magnetic resonance imaging study

To describe the neural substrates of successful episodic long-term memory encoding, we collected functional magnetic-resonance imaging data as participants completed an arbitrary delayed auditory paired-association learning task. During the task, subjects learned predefined but hidden stimulus pairs by trial and error based on visual feedback. Delay period activity represents the retrieval of the relationship between the cue item and its candidate for associates, that is, working memory. Our hypothesis was that the neural substrates of working memory would be related to long-term memory encoding in a performance-dependent manner. Thus, inter-individual variance in performance following a fixed learning set would be associated with differing neural activations during the delay period. The number of learning trials was adjusted such that performance following completion of the learning set varied across subjects. Each trial consisted of the successive presentation of two stimuli (first stimulus and second stimulus [S2]) with a fixed delay interval, allowing extraction of sustained activity during the delay period. Sustained activities during the delay period were found in the bilateral dorsolateral prefrontal cortex, intraparietal sulcus, and left ventrolateral prefrontal cortex, as well as the premotor and pre-supplementary motor areas. The activities did not change in strength across learning, suggesting that these effects represent working memory components. The sustained activity in the ventrolateral prefrontal region was correlated with task performance. Task performance was also positively correlated with the decrement in S2/feedback-related activity during learning in the superior temporal sulcus, a region previously shown to be involved in association learning. These findings are consistent with lesion and neuroimaging studies showing that the ventrolateral prefrontal cortex plays an important role in long-term memory encoding, and raise the possibility that working memory processes interact with long-term memory formation as represented by the covariation of activity in the superior temporal sulcus and the ventrolateral prefrontal cortex.     

健康人大脑和小脑空间记忆认知功能的fMRI研究

本研究应用功能磁共振成像(functional magnetic resonance imaging,fMRI)技术,检测了健康人大脑和小脑参与空间记忆的认知过程。通过对10名右利手健康志愿者进行一项短时空间记忆任务作业的同时进行脑功能磁共振扫描,实验采用组块设计,任务与对照任务交替进行,数据采用SPM99软件进行数据分析和脑功能区定位。结果显示:当统计阈值设定为P<0.0001时,大脑皮层和右侧小脑一起被显著激活;大脑皮层所激活的脑区有双侧顶叶的楔前叶、顶上小叶、缘上回(BA7/40,BA:Brodma-nn Area),双侧前额上、中、下回(BA6/9/47),双侧枕叶和枕颞交界处(BA18/19/37),右侧海马回;左侧中脑黑质及被盖部也被激活。上述结果提示:小脑和大脑皮层一起参与了空间记忆的认知过程。     

Neuronal correlates of encoding and retrieval in episodic memory during a paired-word association learning task: a functional magnetic resonance imaging study

The investigation of memory function using functional magnetic resonance imaging (fMRI) is an expanding field of research. The aim of this study was to demonstrate brain-activity patterns related to a word-pair association task employing a whole-brain EPI sequence. Six right-handed, healthy male volunteers (mean age: 27.5 years) took part in the study. fMRI was performed at a field strength of 1.5 Tesla with 26–32 slices parallel to the AC-PC line, depending on individual brain size. Distributed brain regions were activated in episodic encoding and retrieval with similarities, but also (distinct) differences in activation patterns. Bilateral prefrontal cortical areas were involved when comparing encoding as well as retrieval to the reference condition (nonsense words). Furthermore, activation was observed in cerebellar areas during encoding, and activation in bilateral parietal areas (precuneus and inferior parietal cortex) was differentially more pronounced during retrieval. The activation of left dorsomedial thalamus during retrieval of high imagery-content word-pair associates may point to the role of this structure in episodic retrieval. The direct cognitive subtraction of encoding minus retrieval yielded a differentially larger left prefrontal activation. There was a differentially higher right prefrontal activation during retrieval than during encoding, underlining the proposed right/left asymmetry for episodic memory processes. Received: 10 June 1998 / Accepted: 23 April 1999     

Working memory for order and the parietal cortex: an event-related functional magnetic resonance imaging study

Memory for order information has been tied to the frontal lobes, however, parietal activation is observed in many functional neuroimaging studies. Here we report functional magnetic resonance findings from an event-related experiment involving working memory for order. Five letters were presented for storage, followed after a delay by two probe items. Probe items could be separated by zero to three positions in the memory set and subjects had to indicate whether the items were in the correct order. Analyses indicate that activation in left parietal cortex shows a systematic decrease in activation with increasing probe distance. This finding is consistent with an earlier study in which we suggested that parietal cortical regions mediate the representation of order information via magnitude codes.     

Neural correlates of conceptual object priming in young and older adults: an event-related functional magnetic resonance imaging study

In this event-related functional magnetic resonance imaging study, we investigated age-related differences in brain activity associated with conceptual repetition priming in young and older adults. Participants performed a speeded “living/nonliving” classification task with 3 repetitions of familiar objects. Both young and older adults showed a similar magnitude of behavioral priming to repeated objects and evidenced repetition-related activation reductions in fusiform gyrus, superior occipital, middle, and inferior temporal cortex, and inferior frontal and insula regions. The neural priming effect in young adults was extensive and continued through both the second and third stimulus repetitions, and neural priming in older adults was markedly attenuated and reached floor at the second repetition. In young adults, greater neural priming in multiple brain regions correlated with greater behavioral facilitation and in older adults, only activation reduction in the left inferior frontal correlated with faster behavioral responses. These findings provide evidence for altered neural priming in older adults despite preserved behavioral priming, and suggest the possibility that age-invariant behavioral priming is observed as a result of more sustained neural processing of stimuli in older adults which might be a form of compensatory neural activity.     

Episodic memory in normal aging and Alzheimer disease: Insights from imaging and behavioral studies

Age-related cognitive changes often include difficulties in retrieving memories, particularly those that rely on personal experiences within their temporal and spatial contexts (i.e., episodic memories). This decline may vary depending on the studied phase (i.e., encoding, storage or retrieval), according to inter-individual differences, and whether we are talking about normal or pathological (e.g., Alzheimer disease; AD) aging. Such cognitive changes are associated with different structural and functional alterations in the human neural network that underpins episodic memory. The prefrontal cortex is the first structure to be affected by age, followed by the medial temporal lobe (MTL), the parietal cortex and the cerebellum. In AD, however, the modifications occur mainly in the MTL (hippocampus and adjacent structures) before spreading to the neocortex. In this review, we will present results that attempt to characterize normal and pathological cognitive aging at multiple levels by integrating structural, behavioral, inter-individual and neuroimaging measures of episodic memory.     

Converging functional magnetic resonance imaging evidence for a role of the left inferior frontal lobe in semantic retention during language comprehension

Increasing evidence supports dissociable short-term memory (STM) capacities for semantic and phonological representations. Cognitive neuropsychological data suggest that damage to the left inferior and middle frontal gyri are associated with deficits of semantic STM, while damage to inferior parietal areas is associated with deficits of phonological STM. Patients identified as having semantic STM deficits are also impaired on a number of language comprehension and production paradigms. We used one such comprehension task derived from cognitive neuropsychological data to test predictions with functional magnetic resonance imaging (fMRI) using healthy participants. Using a task that required participants to make semantic anomaly judgements, we found significantly greater activation in areas of the left inferior frontal and middle frontal gyri for phrases that required maintenance of multiple words for eventual integration with a subsequent noun or verb. These data are consistent with our previous patient studies (Hanten & Martin, 2000; R. C. Martin & He, 2004; R. C. Martin & Romani, 1994 Martin, R. C. and Romani, C. 1994. Verbal working memory and sentence comprehension: A multiple-components view. Neuropsychology, 8: 506–523. [Crossref] , [Google Scholar]) that suggest that semantic STM is associated with the left inferior and middle frontal gyri and that deficits of semantic STM have particular consequences for comprehension tasks that require maintenance of several word meanings in unintegrated form.     

Saccular stimulation of the human cortex: A functional magnetic resonance imaging study

Recent imaging studies have reported the projection of semicircular canal signals onto wide regions of the cerebral cortex but little is known about otolith projections onto the cerebral cortex. We used functional magnetic resonance imaging (fMRI) to investigate the activation of the cortex by loud clicks that selectively stimulate the sacculus. Twelve normal volunteers were presented with auditory stimuli via an earphone containing a piezo electric element. High-intensity [maximum volume of 120 dB (SPL)] or low-intensity [maximum volume of 110 dB (SPL)] clicks were delivered at a frequency of 1 Hz and lasted 1 ms. We first checked that the high-intensity, but not low-intensity, clicks stimulated the sacculus by determining the vestibular evoked myogenic potentials. We then analyzed two task conditions (high- and low-intensity clicks) in a boxcar paradigm. We obtained gradient echo echo-planar images by using a 1.5 T MRI system. We analyzed the fMRI time series data with SPM2. High-intensity clicks activated wide areas of the cortex, namely, the frontal lobe (prefrontal cortex, premotor cortex, and frontal eye fields), parietal lobe (the region around the intraparietal sulcus, temporo-parietal junction, and paracentral lobule), and cingulate cortex. These areas are similar to those reported in previous imaging studies that analyzed the cortical responses to the activation of the semicircular canals. Thus, semicircular canal and otolith/saccular signals may be processed in similar regions of the human cortex.     

Negative BOLD during tongue movement: A functional magnetic resonance imaging study

The aim of this functional magnetic resonance imaging (fMRI) study was to evaluate negative blood oxygen level-dependent (BOLD) signals during voluntary tongue movement. Deactivated (Negative BOLD) regions included the posterior parietal cortex (PPC), precuneus, and middle temporal gyrus. Activated (Positive BOLD) regions included the primary somatosensory-motor area (SMI), inferior parietal lobule, medial frontal gyrus, superior temporal gyrus, insula, lentiform nucleus, and thalamus. The results were not consistent with previous studies involving unilateral hand and finger movements showing the deactivation of motor-related cortical areas including the ipsilateral MI. The areas of Negative BOLD in the PPC and precuneus might reflect specific neural networks relating to voluntary tongue movement.     

Light sedation with short habituation time for large-scale functional magnetic resonance imaging studies in rats

Traditionally, preclinical resting state functional magnetic resonance imaging (fMRI) studies have been performed in anesthetized animals. Nevertheless, as anesthesia affects the functional connectivity (FC) in the brain, there has been a growing interest in imaging in the awake state. Obviously, awake imaging requires resource- and time-consuming habituation prior to data acquisition to reduce the stress and motion of the animals. Light sedation has been a less widely exploited alternative for awake imaging, requiring shorter habituation times, while still reducing the effect of anesthesia. Here, we imaged 102 rats under light sedation and 10 awake animals to conduct an FC analysis. We established an automated data-processing pipeline suitable for both groups. Additionally, the same pipeline was used on data obtained from an openly available awake rat database (289 measurements in 90 rats). The FC pattern in the light sedation measurements closely resembled the corresponding patterns in both onsite and offsite awake datasets. However, fewer datasets had to be excluded due to movement in rats with light sedation. The temporal analysis of FC in the lightly sedated group indicated a lingering effect of anesthesia that stabilized after the first 5 min. In summary, our results indicate that the light sedation protocol is a valid alternative for large-scale studies where awake protocols may become prohibitively resource-demanding, as it provides similar results to awake imaging, preserves more scans, and requires shorter habituation times. The large amount of fMRI data obtained in this work are openly available for further analyses.     

Midlife memory ability accounts for brain activity differences in healthy aging

Cross-sectional neuroimaging studies suggest that hippocampal and prefrontal cortex functions underlie individual differences in memory ability in older individuals, but it is unclear how individual differences in cognitive ability in youth contribute to cognitive and neuroimaging measures in older age. Here, we investigated the relative influences of midlife memory ability and age-related memory change on memory-related BOLD-signal variability at one time point, using a sample from a longitudinal population-based aging study (N = 203, aged 55–80 years). Hierarchical regression analyses showed that midlife memory ability, assessed 15–20 years earlier, explained at least as much variance as memory change in clusters in the left inferior prefrontal cortex and the bilateral hippocampus, during memory encoding. Furthermore, memory change estimates demonstrated higher sensitivity than current memory levels in identifying distinct frontal regions where activity was selectively related to age-related memory change, as opposed to midlife memory. These findings highlight challenges in interpreting individual differences in neurocognitive measures as age-related changes in the absence of longitudinal data and also demonstrate the improved sensitivity of longitudinal measures.     

From fetus to older age: A review of brain metabolic changes across the lifespan

IntroductionThe knowledge of metabolic changes across the lifespan is poorly understood. Thus we systematically reviewed the available literature to determine the changes in brain biochemical composition from fetus to older age and tried to explain them in the context of neural, cognitive, and behavioural changes.MethodsThe search identified 1262 articles regarding proton magnetic resonance spectroscopy (1H MRS) examinations through December 2017. The following data was extracted: age range of the subjects, number of subjects studied, brain regions studied, MRS sequence used, echo time, MR system, method of statistical analysis, metabolites analyzed, significant differences in metabolites concentrations with age as well as the way of presentation of the results.Results82 studies that described brain metabolite changes with age were identified. Reports on metabolic changes related to healthy aging were analyzed and discussed among six basic age groups: fetuses, infants, children, adolescents, adults, and the elderly as well as between groups and during the whole lifetime.DiscussionThe results presented in the reviewed papers provide evidence that normal aging is associated with a number of metabolic changes characteristic for every period of life. Therefore, it can be concluded that the age matching is essential for comparative studies of disease states using 1H MRS.     

A functional magnetic resonance imaging study of the effects of pergolide, a dopamine receptor agonist, on component processes of working memory

Working memory is an important cognitive process dependent on a network of prefrontal and posterior cortical regions. In this study we tested the effects of the mixed D1-D2 dopamine receptor agonist pergolide on component processes of human working memory using functional magnetic resonance imaging (fMRI). An event-related trial design allowed separation of the effects on encoding, maintenance, and retrieval processes. Subjects were tested with spatial and object memoranda to investigate modality-specific effects of dopaminergic stimulation. We also measured baseline working memory capacity as previous studies have shown that effects of dopamine agonists vary with working memory span. Pergolide improved reaction time for high-span subjects and impaired reaction time for low-span subjects. This span-dependent change in behavior was accompanied by span-dependent changes in delay-related activity in the premotor cortex. We also found evidence for modality-specific effects of pergolide only during the response period. Pergolide increased activity for spatial memoranda and decreased activity for object memoranda in task-related regions including the prefrontal and parietal cortices.     

特殊感觉障碍人群(盲人)脑功能磁共振成像研究

目的：探讨特殊感觉障碍人群(盲人)大脑枕叶视皮层及相关结构在肓文阅读时充当的角色。方法：采用功能磁共振成像(fMRI)技术研究8名先天和后天盲人，在刺激(触摸中国盲文)和静止等两种对比条件下采集枕叶和有关皮层的回波平面图象。结果：盲人阅读中国盲文时在距状裂附近的纹状区、纹外区包括次级视觉中枢及与视觉信息处理有关的脑区均有激活信号出现，而正常人触摸盲文时上述脑区未出现明显激活信号。结论：枕叶可能参与肓文阅读过程的触觉信息识别：     

Depression symptoms and cognitive control of emotion cues: a functional magnetic resonance imaging study

Few studies have examined associations between depressive symptoms and alterations in neural systems that subserve cognitive control. Cognitive control was assessed with an exogenous cueing task using happy, sad, and neutral facial expressions as cues among women with mild to moderate symptoms of depression and a non-depressed control group while functional magnetic resonance imaging (fMRI) measured brain activity. Amygdala and medial/orbital prefrontal cortex (PFC) response to valid emotion cues did not differ as a function of depression symptoms. However, significant depression group differences were observed when task demands required cognitive control. Participants with elevated depression symptoms showed weaker activation in right and left lateral PFC and parietal regions when shifting attentional focus away from invalid emotion cues. No depression group differences were observed for invalid non-emotional cues. Findings suggest that mild to moderate depression symptoms are associated with altered function in brain regions that mediate cognitive control of emotional information.     

A functional magnetic resonance imaging navigated repetitive transcranial magnetic stimulation study of the posterior parietal cortex in normal pain and hyperalgesia

Noxious stimuli activate a complex cerebral network. During central sensitization to pain, activity in most of these areas is changed. One of these areas is the posterior parietal cortex (PPC). The role of the PPC during processing of acute pain as well as hyperalgesia and tactile allodynia remains elusive. Therefore, we performed a functional magnetic resonance imaging (fMRI) based, neuro-navigated, repetitive transcranial magnetic stimulation (rTMS) study in 10 healthy volunteers. Firstly, pin-prick hyperalgesia was provoked on the right volar forearm, using the model of electrically-induced secondary mechanical hyperalgesia. fMRI was performed during pin-prick stimulation inside and outside the hyperalgesic areas. Secondly, on four different experimental sessions, the left and right individual intraparietal BOLD peak-activations were used as targets for a sham-controlled 1 Hz rTMS paradigm of 10 min duration. We measured psychophysically the (i) electrical pain stimulus intensity on an 11-point numeric pain rating scale (NRS, 0–10), the (ii) area of hyperalgesia, and the (iii) area of dynamic mechanical allodynia. Sham stimulation or rTMS was performed 16 min after induction of pin-prick hyperalgesia and tactile allodynia. Compared to sham stimulation, no significant effect of rTMS was observed on pain stimulus intensity and the area of allodynia. However, a reduction of the hyperalgesic area was observed for rTMS of the left PPC (P<0.05). We discuss the role of the PPC in central sensitization to pain, in spatial discrimination of pain stimuli and in spatial-attention to pain stimuli.     

Sleep stabilizes visuomotor adaptation memory: a functional magnetic resonance imaging study

The beneficial effect of sleep on motor memory consolidation is well known for motor sequence memory, but remains unsettled for visuomotor adaptation in humans. The aim of this study was to characterize more clearly the influence of sleep on consolidation of visuomotor adaptation using a between‐subjects functional magnetic resonance imaging (fMRI) design contrasting sleep to total sleep deprivation. Our behavioural results, based on seven different parameters, show that sleep stabilizes performance whereas sleep deprivation deteriorates it. During training, while a set of cerebellar, striatal and cortical areas is activated in proportion to performance improvement, the recruitment of the hippocampus and frontal cortex protects motor memory against the detrimental effects of sleep deprivation. During retest after sleep loss a cerebello–cortical network, usually involved in the earliest stage of learning, was recruited to perform the task. In contrast, no changes in cerebral activity were observed after sleep, suggesting that it may only support the stabilization of the visuomotor adaptation memory trace.     

Speed and memory in the WAIS-III Digit Symbol--Coding subtest across the adult lifespan.

The primary role of speed in determining Digit Symbol scores is well established. Among the important questions that remain to be resolved are: (1) whether speed accounts for all of the age-related decline in Digit Symbol scores, and (2) whether memory ability makes any significant contribution to Digit Symbol performance, especially after controlling for speed. We analyzed data from the WAIS-III/WMS-III standardization sample to resolve these issues. As expected, speed (Digit Symbol-Copy) correlated very strongly with Digit Symbol--Coding. Memory (Digit Symbol--Incidental Learning or WMS-III index scores) correlated more moderately with Digit Symbol-Coding. Even after controlling for variance in Coding explained by Copying, a statistically significant proportion of the residual variance was explicable in terms of memory functions. The contribution of memory to Digit Symbol--Coding, while relatively small, is real. In addition, a small portion of the age-associated decline in Coding scores cannot be accounted for by Copying scores.     

Differences in functional magnetic resonance imaging of sensorimotor cortex during static and dynamic finger flexion

 Functional magnetic resonance imaging (fMRI) studies of the human motor system have commonly used movement paradigms which contain a dynamic component; however, the relationship between the fMRI signal for motor tasks with and without a dynamic component is not known. We have investigated the relationship between the fMRI signal during a static finger flexion task and during dynamic finger flexion at 1–3 Hz, each at two levels of force (5% and 10% of maximum voluntary contraction). A small fMRI response could be recorded from only a few subjects during the static tasks. In contrast, a substantial fMRI response occurred during dynamic tasks in all subjects at both levels of force. The fMRI response was not significantly correlated with force or movement rate during the dynamic tasks. It is concluded that the factors responsible for generating an fMRI response are fundamentally different during steady contractions compared to those involving a dynamic component, and that the fMRI signal may be more sensitive to changes in the pattern of neural activation rather than the ongoing firing rate or extent of activation. Received: 24 March 1998 / Accepted: 9 October 1998