Imaging cognition II: An empirical review of 275 PET and fMRI studies

Positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) have been extensively used to explore the functional neuroanatomy of cognitive functions. Here we review 275 PET and fMRI studies of attention (sustained, selective, Stroop, orientation, divided), perception (object, face, space/motion, smell), imagery (object, space/motion), language (written/spoken word recognition, spoken/no spoken response), working memory (verbal/numeric, object, spatial, problem solving), semantic memory retrieval (categorization, generation), episodic memory encoding (verbal, object, spatial), episodic memory retrieval (verbal, nonverbal, success, effort, mode, context), priming (perceptual, conceptual), and procedural memory (conditioning, motor, and nonmotor skill learning). To identify consistent activation patterns associated with these cognitive operations, data from 412 contrasts were summarized at the level of cortical Brodmann's areas, insula, thalamus, medial-temporal lobe (including hippocampus), basal ganglia, and cerebellum. For perception and imagery, activation patterns included primary and secondary regions in the dorsal and ventral pathways. For attention and working memory, activations were usually found in prefrontal and parietal regions. For language and semantic memory retrieval, typical regions included left prefrontal and temporal regions. For episodic memory encoding, consistently activated regions included left prefrontal and medial temporal regions. For episodic memory retrieval, activation patterns included prefrontal, medial temporal, and posterior midline regions. For priming, deactivations in prefrontal (conceptual) or extrastriate (perceptual) regions were consistently seen. For procedural memory, activations were found in motor as well as in non-motor brain areas. Analysis of regional activations across cognitive domains suggested that several brain regions, including the cerebellum, are engaged by a variety of cognitive challenges. These observations are discussed in relation to functional specialization as well as functional integration.     

Visualizing how one brain understands another: a PET study of theory of mind

OBJECTIVE: Theory of mind (TOM), or "mentalizing," refers to the ability to attribute mental states to self and others. Inferring what people are thinking and feeling is an important aspect of human social interaction, and it is also an important aspect of both psychiatric diagnosis and treatment. The authors conducted a positron emission tomography (PET) study to examine the neural substrates of TOM, using a task that mimics real-life social interaction. METHOD: Thirteen healthy volunteers underwent [(15)O]H(2)O PET while performing an experimental task and a control task. During the experimental task they created a "story" about the mental state of a stranger whom they imagined encountering on a park bench. During the control task, they read aloud a story requiring no mental state attribution. RESULTS: The TOM task activated an extensive neural network that included the medial frontal cortex, the superior frontal cortex, the anterior and retrosplenial cingulate, and the anterior temporal pole; most of these activations were limited to the left hemisphere. In addition, the largest activation was in the contralateral right cerebellum, as well as the anterior vermis. CONCLUSIONS: A language-based TOM task activated distributed brain regions that are important for representing mental states of the self and others, retrieving memory of personal experiences, and coordinating and monitoring the overall performance of the task. The activations in the medial frontal cortex replicate findings in previous TOM studies, while the activations in the cerebellum reinforce the growing evidence that the cerebellum performs cognitive functions in the human brain.     

Brain regions associated with episodic retrieval in normal aging and Alzheimer's disease.

OBJECTIVE: To examine patterns of brain activation during verbal episodic retrieval in normal elderly subjects and patients in an early phase of AD. BACKGROUND: It is established that 1) a profound episodic memory impairment is a cardinal symptom of AD; and 2) some of the earliest brain changes in this disease occur in regions critical to episodic memory, such as the hippocampus and neighboring regions. Yet, it remains largely unknown whether the episodic memory deficit seen in AD is paralleled by concomitant alterations in brain activity during actual task performance in these or other brain areas. METHODS: Using PET, blood flow was assessed in normal elderly subjects and patients with early AD during two retrieval conditions involving completion of word stems: baseline and cued recall. RESULTS: The patients with AD showed a marked performance deficit in cued recall, although the two groups were indistinguishable in the baseline task condition. Both groups showed bilateral activity in orbital and dorsolateral prefrontal cortex, left precuneus, and right cerebellum, as well as decreased activity in distinct left temporal regions during cued recall. The normal elderly alone activated the left parietal cortex and the left hippocampal formation during episodic retrieval. By contrast, AD-related increases in activity during cued recall were observed in the left orbital prefrontal cortex and left cerebellum. CONCLUSIONS: The similar patterns of activations in the two groups suggest that a large distributed network involved in episodic memory retrieval functions relatively normally in early AD. Those retrieval activations seen in the normal elderly, as opposed to the patients, may reflect AD-related failures in semantic processing and successful recollection of the target information, respectively. Finally, the AD-related increases in activity were interpreted in terms of compensatory reactions to the difficulties in performing the episodic memory task.     

Subtle cognitive deficits after cerebellar infarcts

The role of the cerebellum in cognitive functions has been under debate. We investigated the neuropsychological functioning of patients with cerebellar lesions (infarcts) and evaluated the significance of laterality in cognitive symptoms. Twenty-six patients with exclusive cerebellar lesions as verified by clinical and neuroradiological findings underwent a neuropsychological assessment at the acute stage and at 3 months. Their performance was compared with 14 controls, also assessed twice. The focus was on four domains: visuospatial/motor functions, episodic memory, working memory and attentional shifting/execution. Both groups improved over time. Statistical differences emerged in tests in the visuomotor domain as well as in the episodic and working memory domains. Patients with left cerebellar lesion were slow in a visuospatial task, whereas those with right cerebellar lesions had verbal memory difficulty compared with controls. By 3 months, 77% of the patients had returned to work, and only one had cognitive impairment and did not return to work. Our results indicate that cerebellar infarcts may result in subtle cognitive changes perhaps primarily related to working memory deficit. The symptoms may be mediated by the contralateral cortical hemisphere, left cerebellar infarcts producing mild right hemispheral dysfunction and right cerebellar infarct producing mild left hemispheral dysfunction.     

Attention-related activity during episodic memory retrieval: a cross-function fMRI study

In functional neuroimaging studies of episodic retrieval (ER), activations in prefrontal, parietal, anterior cingulate, and thalamic regions are typically attributed to episodic retrieval processes. However, these activations are also frequent during visual attention (VA) tasks, suggesting that their role in ER may reflect attentional rather than mnemonic processes. To investigate this possibility, we directly compared brain activity during ER and VA tasks using event-related fMRI. The ER task was a word recognition test with a retrieval mode component, and the VA task was a target detection task with a sustained attention component. The study yielded three main findings. First, a common fronto-parietal-cingulate-thalamic network was found for ER and VA, suggesting that the involvement of these regions during ER reflects general attentional processes. This idea is compatible with some of the interpretations proposed in the ER literature (e.g. postretrieval monitoring), which may be rephrased in terms of attentional processes. Second, several subregions were differentially involved in ER versus VA. For example, the frontopolar cortex and the precuneus were more activated for ER than for VA, possibly reflecting retrieval mode and processing of internally generated stimuli, respectively. Finally, the study yielded an unexpected finding: some medial temporal lobe regions were similarly activated for ER and VA. This finding suggests that the medial temporal lobes may be involved in indexing representations within the focus of consciousness, regardless of whether they are mnemonic or perceptual. Overall, the present results suggest that many of the activations attributed to specific cognitive processes, such as episodic memory, may actually reflect more general cognitive operations.     

Common and unique neural activations in autobiographical, episodic, and semantic retrieval

This study sought to explore the neural correlates that underlie autobiographical, episodic, and semantic memory. Autobiographical memory was defined as the conscious recollection of personally relevant events, episodic memory as the recall of stimuli presented in the laboratory, and semantic memory as the retrieval of factual information and general knowledge about the world. Our objective was to delineate common neural activations, reflecting a functional overlap, and unique neural activations, reflecting functional dissociation of these memory processes. We conducted an event-related functional magnetic resonance imaging study in which we utilized the same pictorial stimuli but manipulated retrieval demands to extract autobiographical, episodic, or semantic memories. The results show a functional overlap of the three types of memory retrieval in the inferior frontal gyrus, the middle frontal gyrus, the caudate nucleus, the thalamus, and the lingual gyrus. All memory conditions yielded activation of the left medial-temporal lobe; however, we found a functional dissociation within this region. The anterior and superior areas were active in episodic and semantic retrieval, whereas more posterior and inferior areas were active in autobiographical retrieval. Unique activations for each memory type were also delineated, including medial frontal increases for autobiographical, right middle frontal increases for episodic, and right inferior temporal increases for semantic retrieval. These findings suggest a common neural network underlying all declarative memory retrieval, as well as unique neural contributions reflecting the specific properties of retrieved memories.     

Cerebellar Contribution to Pattern Separation of Human Hippocampal Memory Circuits

The cerebellum is a crucial structure for cognitive function as well as motor control. Benign brain tumors such as schwannomas, meningiomas, and epidermoids tend to occur in the cerebellopontine angle cisterns and may cause compression of the posterior lateral cerebellum near the superior posterior fissure, where the eloquent area for cognitive function was recently identified. The present study examined cognitive impairment in patients with benign cerebellar tumors before and after surgical intervention in order to clarify the functional implications of this region in humans. Patients with cerebellar tumors showed deficits in psychomotor speed and working memory compared with healthy controls. Moreover, these impairments were more pronounced in patients with right cerebellar tumors. Functional magnetic resonance imaging during performance of a lure task also demonstrated that cerebellar tumors affected pattern separation or the ability to distinguish similar experiences of episodic memory or events with discrete, non-overlapping representations, which is one of the important cognitive functions related to the hippocampus. The present findings indicate that compression of the human posterior lateral cerebellum affects hippocampal memory function.     

Common prefrontal activations during working memory,episodic memory,and semantic memory

Regions of the prefrontal cortex (PFC) are typically activated in many different cognitive functions. In most studies, the focus has been on the role of specific PFC regions in specific cognitive domains, but more recently similarities in PFC activations across cognitive domains have been stressed. Such similarities may suggest that a region mediates a common function across a variety of cognitive tasks. In this study, we compared the activation patterns associated with tests of working memory, semantic memory and episodic memory. The results converged on a general involvement of four regions across memory tests. These were located in left frontopolar cortex, left mid-ventrolateral PFC, left mid-dorsolateral PFC and dorsal anterior cingulate cortex. These findings provide evidence that some PFC regions are engaged during many different memory tests. The findings are discussed in relation to theories about the functional contribution of the PFC regions and the architecture of memory.     

Autobiographical and episodic memory--one and the same? Evidence from prefrontal activation in neuroimaging studies

Laboratory investigations of episodic memory often require participants to encode and later retrieve lists of items (words, pictures, or faces). The underlying assumption is that recollection of items from the list is analogous to recollection of events from one's past, i.e. autobiographical re-experiencing. Functional neuroimaging studies of episodic memory have provided extensive evidence suggesting that regions of the prefrontal cortex (PFC) play a role in episodic memory retrieval. A review of PFC activations reported in imaging studies of autobiographical memory and matched sub-sets of list-learning episodic memory studies reveals patterns of similarity but also substantial differences. Episodic memory studies often report activations in the right mid-dorsolateral PFC, but such activations are absent in autobiographical memory studies. Additionally, activations in the ventromedial PFC, primarily on the left, are almost invariably found in autobiographical memory studies, but rarely occur in studies of episodic memory. It is suggested that these two regions mediate different modes of post-retrieval monitoring and verification. Autobiographical memory relies on quick intuitive 'feeling of rightness' to monitor the veracity and cohesiveness of retrieved memories in relation to an activated self-schema. Episodic memory for lists requires more conscious elaborate monitoring to avoid omissions, commissions and repetitions. The present analysis suggests that care and caution should be exercised in extrapolating from the way we recollect 'events' from a list learned in the laboratory to the way we recollect events from our lives.     

Gender differences in the functional neuroanatomy of emotional episodic autobiographical memory

Autobiographical memory is based on interactions between episodic memory contents, associated emotions, and a sense of self-continuity along the time axis of one's life. The functional neuroanatomy subserving autobiographical memory is known to include prefrontal, medial and lateral temporal, as well as retrosplenial brain areas; however, whether gender differences exist in neural correlates of autobiographical memory remains to be clarified. We reanalyzed data from a previous functional magnetic resonance imaging (fMRI) experiment to investigate gender-related differences in the neural bases of autobiographical memories with differential remoteness and emotional valence. On the behavioral level, there were no significant gender differences in memory performance or emotional intensity of memories. Activations common to males and females during autobiographical memory retrieval were observed in a bilateral network of brain areas comprising medial and lateral temporal regions, including hippocampal and parahippocampal structures, posterior cingulate, as well as prefrontal cortex. In males (relative to females), all types of autobiographical memories investigated were associated with differential activation of the left parahippocampal gyrus. By contrast, right dorsolateral prefrontal cortex was activated differentially by females. In addition, the right insula was activated differentially in females during remote and negative memory retrieval. The data show gender-related differential neural activations within the network subserving autobiographical memory in both genders. We suggest that the differential activations may reflect gender-specific cognitive strategies during access to autobiographical memories that do not necessarily affect the behavioral level of memory performance and emotionality.     

Neural bases of learning and memory: functional neuroimaging evidence

Positron emission tomography and functional magnetic resonance imaging studies have identified brain regions associated with different forms of memory. Working memory has been associated primarily with the bilateral prefrontal and parietal regions; semantic memory with the left prefrontal and temporal regions; episodic memory encoding with the left prefrontal and medial temporal regions; episodic memory retrieval with the right prefrontal, posterior midline and medial temporal regions; and skill learning with the motor, parietal, and subcortical regions. Recent studies have provided higher specificity, by dissociating the neural correlates of different subcomponents of complex memory tasks, and the cognitive roles of different subregions of larger brain areas.     

The posterior parietal cortex in recognition memory: a neuropsychological study

Several recent functional neuroimaging studies have reported robust bilateral activation (L>R) in lateral posterior parietal cortex and precuneus during recognition memory retrieval tasks. It has not yet been determined what cognitive processes are represented by those activations. In order to examine whether parietal lobe-based processes are necessary for basic episodic recognition abilities, we tested a group of 17 first-incident CVA patients whose cortical damage included (but was not limited to) extensive unilateral posterior parietal lesions. These patients performed a series of tasks that yielded parietal activations in previous fMRI studies: yes/no recognition judgments on visual words and on colored object pictures and identifiable environmental sounds. We found that patients with left hemisphere lesions were not impaired compared to controls in any of the tasks. Patients with right hemisphere lesions were not significantly impaired in memory for visual words, but were impaired in recognition of object pictures and sounds. Two lesion--behavior analyses--area-based correlations and voxel-based lesion symptom mapping (VLSM)---indicate that these impairments resulted from extra-parietal damage, specifically to frontal and lateral temporal areas. These findings suggest that extensive parietal damage does not impair recognition performance. We suggest that parietal activations recorded during recognition memory tasks might reflect peri-retrieval processes, such as the storage of retrieved memoranda in a working memory buffer for further cognitive processing.     

Frontal and posterior cingulate metabolic impairment in the behavioral variant of frontotemporal dementia with impaired autonoetic consciousness

Although memory dysfunction is not a prominent feature of the behavioral variant of frontotemporal dementia (bv-FTD), there is evidence of specific deficits of episodic memory in these patients. They also have problems monitoring their memory performance. The objective of the present study was to explore the ability to consciously retrieve own encoding of the context of events (autonoetic consciousness) and the ability to monitor memory performance using feeling-of-knowing (FOK) in bv-FTD. Analyses of the patients' cerebral metabolism (FDG-PET) allowed an examination of whether impaired episodic memory in bv-FTD is associated with the frontal dysfunction characteristic of the pathology or a dysfunction of memory-specific regions pertaining to Papez's circuit. Data were obtained from eight bv-FTD patients and 26 healthy controls. Autonoetic consciousness was evaluated by Remember responses during the recognition memory phase of the FOK experiment. As a group, bv-FTD patients demonstrated a decline in autonoetic consciousness and FOK accuracy at the chance level. While memory monitoring was impaired in most (seven) patients, four bv-FTD participants had individual impairment of autonoetic consciousness. They specifically showed reduced metabolism in the anterior medial prefrontal cortex, the left dorsolateral prefrontal cortex (near the superior frontal sulcus), parietal regions, and the posterior cingulate cortex. These findings were tentatively interpreted by considering the role of the metabolically impaired brain regions in self-referential processes, suggesting that the bv-FTD patients' problem consciously retrieving episodic memories may stem at least partly from deficient access to and maintenance/use of information about the self. Frontal and posterior cingulate metabolic impairment in the behavioral variant of frontotemporal dementia with impaired autonoetic consciousness     

Early parietal response in episodic retrieval revealed with MEG

Recent neuroimaging and lesion studies have led to competing hypotheses for potential roles of the left lateral parietal lobe in episodic memory retrieval. These hypotheses may be dissociated by whether they imply a role in preretrieval or postretrieval processes. For example, one hypothesis is the left parietal cortex (particularly in more ventral subregions) forms part of an “episodic buffer” that supports the online representation of the retrieved target, a role that is, by definition, postretrieval. An alternate view maintains parietal activity (particularly in more dorsal subregions) contributes to top‐down orientation of attention to retrieval search, a preretrieval role. The present investigation seeks to reveal the earliest onset of lateral parietal activity in three anatomically‐defined subregions of the left lateral parietal cortex to identify any preretrieval activation. Subjects performed a pair‐cued recall task while neural activity was recorded with magnetoencephalography (MEG) at millisecond temporal resolution. MEG data were then mapped to each subject's cortical surface using dynamic statistical parametric mapping (dSPM). Both dorsal and ventral regions showed retrieval‐related activations beginning within ～100 ms of the cue to retrieve and lasting up to 400 ms. We conclude that this early and transient pattern of activity in lateral parietal cortex is most consistent with a preretrieval role, possibly in directing attention to episodic memory retrieval. Hum Brain Mapp, 2011. © 2010 Wiley‐Liss, Inc.     

Sustained and transient neural modulations in prefrontal cortex related to declarative long-term memory, working memory, and attention

Common activations in prefrontal cortex (PFC) during episodic and semantic long-term memory (LTM) tasks have been hypothesized to reflect functional overlap in terms of working memory (WM) and cognitive control. To evaluate a WM account of LTM-general activations, the present study took into consideration that cognitive task performance depends on the dynamic operation of multiple component processes, some of which are stimulus-synchronous and transient in nature; and some that are engaged throughout a task in a sustained fashion. PFC and WM may be implicated in both of these temporally independent components. To elucidate these possibilities we employed mixed blocked/event-related functional magnetic resonance imaging (fMRI) procedures to assess the extent to which sustained or transient activation patterns overlapped across tasks indexing episodic and semantic LTM, attention (ATT), and WM. Within PFC, ventrolateral and medial areas exhibited sustained activity across all tasks, whereas more anterior regions including right frontopolar cortex were commonly engaged in sustained processing during the three memory tasks. These findings do not support a WM account of sustained frontal responses during LTM tasks, but instead suggest that the pattern that was common to all tasks reflects general attentional set/vigilance, and that the shared WM-LTM pattern mediates control processes related to upholding task set. Transient responses during the three memory tasks were assessed relative to ATT to isolate item-specific mnemonic processes and were found to be largely distinct from sustained effects. Task-specific effects were observed for each memory task. In addition, a common item response for all memory tasks involved left dorsolateral PFC (DLPFC). The latter response might be seen as reflecting WM processes during LTM retrieval. Thus, our findings suggest that a WM account of shared PFC recruitment in LTM tasks holds for common transient item-related responses rather than sustained state-related responses that are better seen as reflecting more general attentional/control processes.     

Associations evoked during memory encoding recruit the context-network

An extensive cortical network consisting of structures in the medial temporal lobe (hippocampus and parahippocampal cortex), lateral parietal cortex, retrosplenial cortex, and medial prefrontal cortex has recently attracted attention in cognitive neuroscience research, linking the network to both episodic memory and spatial processing. It has been suggested that its function may be best characterized as supporting the processing of contextual associations (context network). In this study, we explored whether the role of this network in contextual processing extends to associations that are evoked in a spontaneous manner. In a novel memory encoding task, participants indicated whether they encoded pictures (objects and novel faces) based on an evoked association or based on a perceptual feature. Memory encoding with subjective associations enhanced memory formation relative to feature-based encoding, and this effect was more pronounced for rapidly evoked associations. Functional magnetic resonance imaging during encoding yielded significant activations in all regions of the context network, i.e., medial prefrontal cortex, lateral parietal cortex, retrosplenial cortex, and posterior medial temporal lobe for the associative vs. feature-based comparisons. The low number of misses did not permit the analysis of a subsequent memory contrast. Our data suggest that the context network, which includes the posterior hippocampus and parahippocampal cortex, might support the linkage of external stimuli to long-term memory representations.     

C-reactive protein is related to memory and medial temporal brain volume in older adults

Recent research suggests a central role for inflammatory mechanisms in cognitive decline that may occur prior to evidence of neurodegeneration. Limited information exists, however, regarding the relationship between low-grade inflammation and cognitive function in healthy older adults. This study examined the relation between inflammation, verbal memory consolidation, and medial temporal lobe volumes in a cohort of older community-dwelling subjects. Subjects included 141 functionally intact, community-dwelling older adults with detectable (n = 76) and undetectable (n = 65) levels of C-reactive protein. A verbal episodic memory measure was administered to all subjects, and measures of delayed recall and recognition memory were assessed. A semiautomated parcellation program was used to analyze structural MRI scans. On the episodic memory task, analysis of covariance revealed a significant CRP group by memory recall interaction, such that participants with detectable levels of CRP evidenced worse performance after a delay compared to those with undetectable levels of CRP. Individuals with detectable CRP also demonstrated lower performance on a measure of recognition memory. Imaging data demonstrated smaller left medial temporal lobe volumes in the detectable CRP group as compared with the undetectable CRP group. These findings underscore a potential role for inflammation in cognitive aging as a modifiable risk factor.     

Alteration of functional neuroanatomy of simple object memory in medial temporal lobe epilepsy patients

Using H2 15O PET, we examined the neuroanatomy associated with a simple form of episodic memory in patients with right or left medial temporal lobe epilepsy and normal healthy controls. When line drawings of common objects were memorized and tested after a 30 min delay, no behavioral difference was found between the patient groups and the controls. However, the patients with epilepsy showed greater cortical activations than the control group on the side ipsilateral to the epileptic focus. rCBF in the anterior thalamic region was enhanced in patients relative to the control group. The results showed that long-term dysfunction of the medial temporal lobe might reinforce alternative memory pathways and recruit a distributed cortical network ipsilateral to their epilepsy focus.     

Common pathway in the medial temporal lobe for storage and recovery of words as revealed by event-related functional MRI

Lesion studies have provided compelling evidence that episodic memory is dependent on the integrity of the medial temporal lobe (MTL). This role of the MTL in episodic memory has been supported by several neuroimaging studies during both episodic encoding and retrieval. After two meta-analyses of positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) studies, we investigated a possible dissociation within the MTL memory system in relation to encoding and retrieval processes. Based on previous reports that specifically related the function of the MTL in episodic memory to successful encoding and actual recovery of information, we applied event-related fMRI to compare successful encoding of words (ES) directly with successful recognition of those same words (RS). Our results did not indicate a clear dissociation between encoding and retrieval activations in the MTL. Instead, a region in the left MTL, covering the parahippocampal cortex and hippocampal formation, which was activated during ES almost completely overlapped with the area that was activated during RS. An additional region in the left anterior MTL, including the entorhinal cortex, was found to be activated exclusively during ES. Research has indicated that a large percentage of cells in this region are particularly sensitive to the relative novelty of stimuli. Our results, therefore, suggest that the parahippocampal/hippocampal region is involved in the formation and subsequent reactivation of memory traces, whereas the activity observed in the entorhinal cortex may reflect elementary memory processes related to novelty detection.     

Loss of memory for auditory-spatial associations following unilateral medial temporal-lobe damage

The goal of the present experiment was to determine the role of medial temporal-lobe structures in episodic memory of auditory-spatial associations. By using a two-alternative forced choice paradigm in which an association between eight different sounds and their spatial location must be recognized, learning abilities over 10 learning sessions were tested in 19 patients who had undergone a right or a left medial temporal-lobe resection for the relief of intractable seizures as well as in nine normal control participants. The data demonstrated that significant learning took place over the successive sessions for all the participants. In addition, the results showed that patients with left but not right medial temporal-lobe lesion were impaired in this learning task as compared to normal participants, suggesting the predominant implication of left medial temporal-lobe structures in auditory-spatial associative learning. The predominant role of left hemisphere structures in this memory task could be explained by a spatial categorical coding, which was enhanced by the use of eight loud-speakers. This result also suggests that the ability to store an episodic event associated with a rich spatial (or temporal) context depends on the left medial temporal-lobe structures. Thus, this finding provides an interesting parallel with data obtained in the visual modality by documenting for the first time the role of the left medial temporal-lobe in episodic learning of auditory-spatial associations.