Brain, Vol. 124, No. 6, 1156-1170,
June 2001
© 2001 Oxford University Press
The effects of bilateral hippocampal damage on fMRI regional activations and interactions during memory retrieval
1 Wellcome Department of Cognitive Neurology, Institute of Neurology, University College London and 2 Developmental Cognitive Neuroscience Unit, Institute of Child Health, University College London, UK and 3 Laboratory of Neuropsychology, National Institute of Mental Health, Bethesda, Maryland, USA
Correspondence to:
Dr Eleanor Maguire, Wellcome Department of Cognitive Neurology, Institute of Neurology, University College London, 12 Queen Square, London WC1N 3BG, UK E-mail: e.maguire{at}fil.ion.ucl.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Using functional magnetic resonance imaging (fMRI) we examined successful retrieval of real-world memories in a patient (Jon) with selective bilateral hippocampal pathology resulting from perinatal hypoxia compared with healthy control subjects. Jon activated the same brain regions during memory retrieval as control subjects, both medial and lateral on the left. In contrast to controls, Jon also activated many homologous regions on the right. In spite of having 50% volume loss bilaterally in his hippocampi, retrieval in Jon was associated with increased activation of the hippocampi. Furthermore, hippocampal activity, as with the controls, was differential, being most responsive to retrieval of autobiographical events compared with other memory types (autobiographical facts, public events, general knowledge). Jon made a distinction between events that the control subjects did not make, namely that some of the autobiographical and public events he clearly remembered, while others he found that he knew about but did not truly remember. His hippocampi and medial frontal cortex were significantly more active during retrieval of events for which he had clear and conscious recollection compared with those he knew as much about, including the context, but could not remember experiencing. Although Jon activates the same network of brain regions as the controls (albeit bilaterally), and with the same pattern of response in the hippocampus, the communication between regions differs from controls with regard to hippocampal–cortical connectivity. In controls there was increased effective connectivity between parahippocampal cortex and hippocampus, specifically during the retrieval of autobiographical events. In contrast, this increase was not apparent in Jon; rather, retrieval of autobiographical events elicited greater interaction between the hippocampus and retrosplenial cortex, and also increased interaction between retrosplenial and medial frontal cortex. This study underlines the value of scanning patients using fMRI while they undertake tasks they can perform, in this case allowing us to confirm the functionality of remaining tissue in the damaged hippocampi, and to appreciate the neural basis of a distinction (remember/know) that control subjects do not make. Besides refining our knowledge of the hippocampal role in autobiographical event memory, this study indicates that recruitment of bilateral regions during memory retrieval, and altered patterns of effective connectivity between brain regions may be important indicators of disordered memory.
autobiographical memory; effective connectivity; fMRI; hippocampus; hypoxia
BA = Brodmann area; SPM = statistical parametric map; T/F = true or false
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
The human hippocampus is known to be important for episodic memory, namely the memory for time- and place-specific experiences, or autobiographical events, that characterize our personal past (Scoville and Milner, 1957
; O'Keefe and Nadel, 1978; Tulving, 1983). It is believed by some that the hippocampus is involved in the memory for explicit information per se, which includes autobiographical events, but also memory for semantic knowledge (Squire, 1992). However, patients with medial temporal lobe damage have been documented who have severely impaired retrograde memory for autobiographical events but relatively preserved semantic knowledge [e.g. McCarthy and Warrington, 1992; Hirano and Noguchi, 1998; Kitchener et al., 1998 (in this case anterograde memory)]. Reports of patients with selective bilateral hippocampal damage who show a similar pattern of impairment and preservation suggest that the hippocampus in particular may be differentially involved in the memory for autobiographical events (Vargha-Khadem et al., 1997; Gadian et al., 2000).
Notwithstanding the devastating consequences of losing autobiographical memory, functional neuroimaging in healthy control subjects has less commonly investigated the neural basis of this type of real-world memory, more often focusing on commonly used laboratory-type tests such as word list learning (e.g. Nyberg et al., 1996; Henson et al., 1999; Rugg et al., 1999). Despite the acknowledged memory impairments in patients with hippocampal damage, activation of the hippocampus in neuroimaging studies of this type is inconsistent and often absent. It is of interest to note that the correlation between commonly used clinical memory tests and the everyday memory problems reported by patients and close observers has been found to be low (Kapur and Pearson, 1983). In contrast to the clinical memory tests, neuroimaging studies that have examined real-world autobiographical event memory do report activation of the hippocampus (Fink et al., 1996; Maguire and Mummery, 1999; Maguire et al., 2000).
In parallel PET (Maguire and Mummery, 1999) and fMRI (functional MRI) (Maguire et al., 2000) studies, the retrieval of four different real-world memory types was systematically examined: memory for personally relevant or autobiographical information, including specific events as outlined above, or facts about oneself not tied to any specific episode. The other two memory types involved information not personally relevant to subjects, again either with a specific spatiotemporal context—public events—or general knowledge not associated with any particular time and place. A largely medial and left-lateralized (see also Conway et al., 1999) network of brain regions was found to support the retrieval of all of the memory types; this included medial frontal cortex, left temporal pole, left hippocampus, left parahippocampal gyrus, left middle temporal gyrus, left temporoparietal junction and retrosplenial cortex. When direct comparisons were made between memory types, however, retrieval of autobiographical event memory compared with any of the other memory types was associated with increased activation of the hippocampus (Maguire and Mummery, 1999), thus mirroring the finding in patients with impairment of autobiographical memory following circumscribed insult to the hippocampi (Vargha-Khadem et al., 1997; Gadian et al., 2000).
Having outlined in control subjects the basic memory retrieval network and differential responses of certain regions therein, one might consider using functional neuroimaging to examine patients with lesions to specific parts of this network. Being able to assess the functionality of damaged brain tissue would be potentially very useful in the clinical context. For example, in cases of hippocampal-damaged patients there is a general assumption in the literature, underpinning their neuropsychological testing, that they are effectively ahippocampal (e.g. Rosenbaum et al., 2000). However, conventional methods—such as neuropsychological testing or structural neuroimaging—cannot determine if residual tissue is still functionally responsive and perhaps contributing to ongoing cognitive operations. While there are PET studies of (non-psychogenic) amnesic patients in the literature (e.g. Reed et al., 1999), these are typically concerned with resting blood flow, without performance of cognitive tasks. In one study where cognitive tasks were used, the patient had suffered a traumatic head injury with cortical and white matter damage (Levine et al., 1998). There is a limited functional neuroimaging literature on progressive conditions (e.g. dementia) and epilepsy (e.g. Swartz and Mandelkern, 1999). However, our concern here is with non-progressive selective hippocampal damage, and there are no reports in the literature of functional neuroimaging during memory testing of patients with such damage. There are reasons for the dearth of such reports. Where patients have difficulty with or simply cannot perform a task, then the interpretation of brain activations is difficult because activation differences may be due to disparities in retrieval success or in the presence or absence of memory traces, etc. compared with control subjects. However, in situations where patients can still perform tasks or a subset of the tasks with accuracy and reaction times similar to that of control subjects, then data can be more usefully considered (for full discussion of this issue, see Price and Friston, 1999).
Recently we had the opportunity to study a patient using fMRI where just such conditions prevailed. Jon, a young man with selective bilateral hippocampal pathology induced by hypoxic–ischaemic episodes of perinatal origin, is one of the cases (Case 2) reported by Vargha-Khadem et al. (1997), with relatively preserved semantic memory but severely impaired episodic memory. On current testing, Jon has a striking episodic memory impairment in the context of excellent general knowledge. He is, however, able to recall some, although for a man aged 22 years, not many, episodic events, both autobiographical and public. His mother also keeps a diary of events that have happened throughout Jon's life. Thus, having interviewed Jon about what he could remember, we were able to check the accuracy of his memory with his parents (see Methods). A great many of the events from his life Jon could not remember, but he recollected just enough episodic events in detail for viable functional scanning. His excellent memory for semantic information was also suitable for scanning in a similar manner to control subjects.
Interestingly, Jon spontaneously made a distinction between events that the control subjects did not make, namely that some of the autobiographical and public events that he recalled he clearly remembered happening, while others he found he knew a great deal about but felt he did not truly recall the event occurring. To the interviewer, there was no difference in the amount or type of information he recalled between the two types. Control subjects remembered all of the events that they recollected. `Remember' judgements are generally associated with the evocation of a specific, previously experienced episode, while `knowing' is typically thought to entail a sense of familiarity, but no information about the source of familiarity (Tulving, 1985; Gardiner and Java, 1991; Knowlton and Squire, 1995). In the case of Jon, he remembers the contexts both for events he remembers and those that he knows, although he clearly has autonoetic conscious awareness (Tulving, 1985) only for the former. Little is definitively known about the neural basis of `remember' and `know' judgements in humans. Some argue for differential involvement of the hippocampus in remember as compared with familiarity judgements (e.g. Aggleton and Brown, 1999; Eldridge et al., 2000), while others either find no hippocampal involvement (Henson et al., 1999) or report involvement of the hippocampus in recognition/familiarity as well (e.g. Manns and Squire, 1999). Of note, there have been no remember/know neuroimaging studies that have used real-world stimuli. In the real-word context, the two undoubtedly interact and cue one another. It is therefore interesting to have what is a rare case where this distinction can be made so clearly with respect to real events.
In summary, in the present study we used fMRI to examine the neural basis of retrieval of real-world memories in a patient with selective bilateral hippocampal damage, where behavioural performance was comparable with that of healthy control subjects. We wished to examine whether the memory retrieval network activated in Jon is the same as that activated in the control subjects, particularly in relation to the recollection of autobiographical events. Differential increase in left hippocampal activity was previously noted in control subjects to be associated with autobiographical events. Given that this is the very structure that is damaged in Jon, would his remaining hippocampal tissue still be functionally responsive, and in a similar manner to control subjects? Furthermore, is there any neural basis to the distinction Jon makes between those events he truly remembers and those he merely knows, matched as they were for amount and type of details recollected? Given that Jon's pathology occurred early in life, is there any neuroimaging evidence of functional reorganization reflecting the greater plasticity of the immature brain? Finally, memory is not just the property of brain regions operating in isolation, but rather of brain networks. We sought to examine if Jon's memory performance, although similar to that of controls, is underpinned by different patterns of interactions between brain regions in the memory retrieval network to those identified in control subjects (Maguire et al., 2000).
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Subjects
Control subjects and Jon (and his mother) gave informed written consent to participate in the study, with approval by the joint National Hospital for Neurology and Neurosurgery/Institute of Neurology Medical Ethics Committee, and (Jon) by the Great Ormond Street Hospital for Children/Institute of Child Health Research Ethics Committee.
Controls
Six healthy right-handed subjects participated in the experiment; four females and two males, age range 28–33 years (mean age 30.5 years).
Patient Jon
Full details of Jon's medical and neuropsychological history and status are described elsewhere (Vargha-Khadem et al., 1997; Gadian et al., 2000). Briefly, this 22-year-old right-handed man was born prematurely after 26 weeks of gestation. A co-twin died shortly after birth. His initial course was characterized by apnoeic attacks which required periods of intubation and ventilation, but thereafter he improved steadily. At 3 years 10 months he had an unconfirmed convulsive episode associated with a cold and cough. While noted to be somewhat clumsy, he developed no motor abnormalities, and no hard neurological signs were/are apparent. At about age 5 years his parents first noticed he had memory difficulties. As described by Vargha-Khadem et al. (1997), three main areas of memory difficulty were noted: poor orientation in date and time; inability to provide reliable accounts of the day's activities, remember telephone conversations, give messages, remember stories or television programmes; and inability to reliably find his way around in familiar surroundings or remember the location of belongings. Despite these problems with episodic memory, he attended mainstream school, achieving near-normal levels of general knowledge (semantic memory).
His most recent evaluation (at age 19 years) using the WAIS-R (Weschler Adult Intelligence Scale—revised) showed Jon to have a VIQ (verbal IQ) of 108, and PIQ (performance IQ) of 120. Detailed neuropsychological test scores are provided in Gadian et al. (2000), and confirm a pattern of severe impairment of episodic memory with relative preservation of semantic memory.
To assess neuropathology in Jon, four quantitative magnetic resonance techniques were used: volumetric measurements, T2 relaxometry, proton magnetic resonance spectroscopy and voxel-based morphometry. Previous findings are described in Vargha-Khadem et al. (1997) and Gadian et al. (2000). Overall, they confirm that Jon's hippocampi are bilaterally shrunken by 50% along their entire length, and that the temporal lobe tissue surrounding Jon's hippocampi appears uncompromised. Voxel-based morphometry data (group data: Jon with four other similar cases) confirmed the reduced grey matter density in the hippocampi bilaterally, revealing (after whole brain evaluation) only one other area of reliable reduction in grey matter, namely the putamen bilaterally. Two other areas of abnormality identified with less confidence were in the ventral thalamus and midbrain (Gadian et al., 2000).
Jon's memory impairments were diagnosed as resulting from selective bilateral hippocampal pathology induced by hypoxic–ischaemic episodes of perinatal origin. Other similar cases (Vargha-Khadem et al., 1997) were either too young, had other complicating medical conditions (e.g. epilepsy), or were not available for testing for the current fMRI experiment.
Design and procedure
Control subjects
Several weeks prior to scanning (mean 5 weeks, range 4–6 weeks), subjects completed a questionnaire and were interviewed in-depth to ascertain details of personal memories and knowledge of public events over a 20-year time period, as well as general knowledge. Subjects were naïve to the purpose of the questions. Statements were constructed from this information for auditory presentation during subsequent scanning and were tailored to individual subject's memories for each of the active conditions. Memories either had a specific locus in time or not, or were relevant to a subject personally or not. This manipulation yielded four memory subtypes: autobiographical events, public events, autobiographical facts and general knowledge.
Both autobiographical events and public events are types of episodic memory, i.e. with a specific spatiotemporal context. This is distinct from factual information, which may pertain specifically to the self or to the world in general, evoking no particular sense of time or place. When debriefed following scanning (as described below), these distinctions were clearly confirmed by the subjects. For example, they were very clear that the public events elicited a sense of the original and specific event, and not merely some factual knowledge about the people or places involved. They also clearly distinguished between events involving themselves, and facts about themselves with no particular spatiotemporal context. Only those memories that subjects recalled in rich detail and for which they declared having a full and conscious recollection of the actual episode (i.e. not re-told to them by a third party) were included in the episodic memory conditions during scanning. Retrieval effort and success were thus similar both within and across memory types. Both autobiographical and public event memories ranged from those that had occurred within a couple of weeks of interview, to remote memories from >20 years ago. Recent and remote memories did not differ in the amount or richness of detail.
During scanning, subjects listened via headphones with their eyes closed and indicated whether each statement was true or false with a keypress response. The inclusion of false stimuli was felt necessary to sustain subjects' attention. The ratio of true : false was 3 : 1. The `false' statements, however, were not completely alien to the subjects; rather, they comprised obvious adjustments to genuine memories. For example, if an actual memory was `You were at Tim's wedding in London', then a false version of this would be `You were at Tim's wedding in Dublin'. We conducted pilot studies to assess what subjects were thinking about during such `false' statements. Because the statements were almost true, subjects (both pilot and the scanned ones) typically report that they thought about the original event as they did with the true statements. All conditions were matched for the number and type of false statements and the position of a false statement within the epoch was counterbalanced across epochs. There was also a control condition during which subjects heard sets of function words (including prepositions and conjunctions) used in the active scans, but which contained no memory information. Subjects decided whether the last word in each control set of function words contained one syllable or not.
Stimuli were matched across epochs for number of syllables. For each statement there was an 8 s time scale; this included presentation of the statement (typically 3–4 s) followed by the subject's response. Four statements were presented per 32 s epoch, and four sets of function words were presented per epoch for control task. Within each epoch, memories from the last 5, 10, 15 and 20 years were included, ensuring that there were no biases in the ages of the memories across the temporally specific conditions (for fMRI direct examination of age of memory effects, see Maguire et al., 2001). Task order was counterbalanced across subjects. The performance of each active task was interspersed with a rest condition. Each memory task was performed four times in total. Examples of each stimulus type: autobiographical event, `You did a tour of Concorde at Heathrow'; autobiographical fact, `Your youngest brother is called Nicolas'; public event, `Windsor Castle was damaged by a fire'; general knowledge, `The people of Holland are known as Dutch'; and control, `He ago always this off otherwise'.
Subjects were debriefed after scanning and did not suggest any task was more difficult than another. All subjects reported that the hearing of a statement evoked a sense of the original event in the case of both autobiographical event and public event memories. When directly probed about the prior interview, none of the subjects reported recalling this during scanning. Even if the interview had been recalled during scanning, it is unlikely that this would have caused the differential effects across the memory types that were in fact found (see below). There was no difference in reaction times between memory types, true or `false' (T/F) statements, or recent and remote events, nor did performance differ between memory types. While the tasks could be described as involving recognition memory, the memories were originally elicited via recall, and subjects report the statements to elicit the memory of the original events. In cases of real-world complex stimuli such as these, the tasks probably reflect a mixture of recognition and recall, so we refer to the task in generic terms as tapping memory `retrieval'.
Jon
The same procedure was employed for patient Jon. As described in the Introduction, the accuracy of his temporally specific memories was verified by additional consultations with his parents. The time scale of such memories was 15 years, shorter than the control subjects given his younger age. Because the time span of memories sampled was the same, i.e. from early childhood to the present, and all subjects were young adults, the age difference was not considered relevant. The same scanning parameters as those used for the control subjects pertained for Jon with the exception that he performed two extra memory conditions, because of the distinction he made between autobiographical and public events that he either clearly remembered or those he merely knew about.
Of note, when he arrived for scanning Jon denied any knowledge of having previously completed the questionnaire or being interviewed. After scanning he expressed surprise at how the experimenter could have known so much about his life. Given this lack of explicit memory for the events prior to scanning, it makes it very unlikely that activations observed during scanning were due to remembering these rather than the memory events of interest. Indeed, in common with the control subjects, Jon confirmed when debriefed that the hearing of a statement evoked a sense of the original episode in the case of both autobiographical and public events that he remembered. As with the control subjects, Jon's performance on the T/F responses during scanning was >88% correct, with no significant difference in performance between memory types. Reaction times did not differ significantly between Jon and control subjects.
Scanning and data analysis
Data were acquired using a 2 T Magnetom VISION (Siemens, Erlangen, Germany) whole body MRI system equipped with a head volume coil. Contiguous multislice T2*-weighted fMRI images were obtained [echo time (TE) = 40 ms] with echo-planar imaging using axial slice orientation. The volume acquired was the whole head (32 slices, each 4 mm thick, 3.2 s per volume). Images were processed and analysed using Statistical Parametric Mapping (SPM99; Wellcome Department of Cognitive Neurology, London, UK). All volumes were realigned to the first volume. A mean image was created using realigned volumes. A structural MRI acquired using a standard three-dimensional T1-weighted sequence (1.5 x 1.5 x 1.5 mm voxel size) was co-registered to the mean (T2*) image. All the images were spatially normalized (Friston et al., 1995; Talairach and Tournoux, 1988) to an echo-planar image template in the same stereotactic space as the Montreal Neurological Institute template (Evans et al., 1993). To reduce confounds due to individual differences in gyral anatomy, the images were smoothed using a Gaussian kernel of 8 mm full-width half maximum. Condition-specific effects were estimated using the general linear model and the theory of Gaussian fields as implemented in SPM99. An adaptive high-pass filter was added to the confound partition of the design matrix to account for low-frequency drifts (Holmes et al., 1997), and the global means were normalized by proportional scaling. Voxel time-series were temporally smoothed with a Gaussian filter (full-width half maximum 6 s). A fixed response box-car model was used to characterize activation effects. Specific effects were tested with appropriate linear contrasts of the parameter estimates for each condition, resulting in a t-statistic for each and every voxel. These t-statistics were transformed to standardized Z-scores and constitute a statistical parametric map (SPM), thresholds for which were set at P < 0.001 (uncorrected) for areas of interest (see Introduction), and P < 0.05 (corrected) for other areas.
Effective connectivity
Full details of this technique as applied to functional neuroimaging data are described elsewhere (Friston et al., 1993; Büchel and Friston, 1997; Köhler et al., 1998; McIntosh et al., 1998; Büchel et al., 1999; Fletcher et al., 1999; Horwitz et al., 1999). In line with previous studies, regions of interest were selected on the basis of P < 0.05 (corrected) significance in the SPM F-map of the group analysis. The activity from the voxel of peak activation in the regions of interest was included in the analysis of effective connectivity. Direct connections between regions within the model were unidirectional to ensure robust estimates. Connections were based on known human and primate anatomy (Van Hoesen, 1982, 1997; Amaral and Insausti, 1990; Arnold et al., 1994). Each connection was assigned a numeric weight, or path coefficient, specific to each experimental condition. These path coefficients are estimates of effective connectivity and represent the response, in units of standard deviation, of the dependent variable (activity in the target region) for a change of 1 SD of the explanatory variable (activity in the source region), while activity elsewhere is held constant.
Two models were constructed. In the first of these, the `null model', the estimates of a path coefficient, or connection strength, were constrained to be equal in two different memory conditions. The alternative model allowed the path coefficient to change as a function of memory type. The significance of the difference between the two models was expressed by the difference in the 
2 goodness of fit indicator (2 difference test or likelihood ratio test; Bollen, 1989). If the model produced a significantly better fit when the path was allowed to change, then this pathway was considered to be affected by memory type.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Aspects of the data from the control subjects have been reported previously (Maguire et al., 2000
) and those presently germane are reprised in the first paragraph below. We then report additional analyses of the control data (see Task-specific effects: control subjects), findings from patient Jon, and direct comparisons between Jon and the control subjects.
Memory retrieval network
Control subjects
Group data for the main comparison of memory [the four memory types (autobiographical events, public events, autobiographical facts, general knowledge) combined] versus the control task shows the brain regions comprising the basic memory retrieval network, irrespective of memory type (Maguire et al., 2000). This comparison revealed a predominantly medial and left-lateralized network consisting of the following regions: medial frontal cortex, temporal pole, hippocampus, anterior middle temporal gyrus, parahippocampal gyrus, retrosplenial cortex and temporoparietal junction (details on the left side of Table 1, and for anatomical clarity displayed in Fig. 1A). The right temporal pole was also activated. One sub-cortical region was activated—the right posterior cerebellum. Comparison of each memory task individually with the control task revealed activation of the identical cortical and limbic network of regions. Subject-specific tests showed that all six subjects showed the same memory network (P < 0.01) (see also the conjunction analysis below). These results accord closely with a similar PET study previously reported in healthy male subjects (Maguire and Mummery, 1999).
|
|
Patient Jon
As with the control subjects, the main comparison of memory (all memory types combined) versus the control task shows the brain regions comprising the basic memory retrieval network in Jon, irrespective of memory type. As evident from the middle of Table 1
, and shown in Fig. 1B, the same set of brain regions was active in Jon as in the control subjects (including the hippocampus) but, in contrast, they were active bilaterally in Jon (right of Table 1). Several additional regions not apparent in the control subjects were also active in Jon (bottom middle of Table 1): right inferior frontal gyrus [Brodmann area (BA 44/45)], right superior frontal gyrus (BA 10), and left middle frontal gyrus (BA 10). Comparison of the individual memory tasks with the control task in Jon showed the same network in each case.
Direct comparisons of the control subjects with Jon
Direct statistical comparisons between Jon and the control subjects were also performed. For the sake of economy we report comparisons pertaining to the memory-control task contrast, as this is fully representative of the individual task comparisons with the control task. First we looked for brain regions which each of the individual control subjects and Jon activate in common—this is called the conjunction. As detailed in Table 2, this showed that each of the subjects including Jon activated the medial and mostly left-lateralized brain regions during memory retrieval. Interestingly, the right hippocampus was also evident in this analysis, but was subthreshold in the group analysis of control subjects.
|
Having noted the commonalities, we next examined the differences between Jon and the control subjects [essentially the interaction: (control subjects Mem-Con)–(Jon Mem-Con)]. The areas more active in the control subjects are shown in Table 3
. Of particular note, the medial temporal region, particularly the hippocampus, is not more active in the control subjects than in Jon. Given the bilateral nature of the memory network in Jon, unsurprisingly he activated many more right-sided regions than the control subjects (see Table 4). Comparison of individual control subjects with Jon in this manner yielded similar results.
|
|
Task-specific effects
Control subjects
When direct comparisons were made between the memory types, autobiographical event memory compared with any of the other memory types showed greater activation of the left hippocampus and medial frontal cortex. For example, even when compared with similarly personal memories but those without a spatiotemporal context, i.e. autobiographical facts, there was greater activation of left hippocampus (–21, –12, –24; Z = 4.07) and medial frontal cortex (BA 10: 0, 57, 12; Z = 6.09) for autobiographical events. In contrast, the other type of episodic memory not personally relevant to the subjects, namely public event memory, showed greater activation of the right middle frontal gyrus (BA 6/8: 36, 15, 51; Z = 6.58) when compared with general knowledge, but no medial temporal changes.
Patient Jon
As with the control subjects, comparisons between the memory types were also performed. In the case of Jon, there was the added distinction he made for the two episodic memory tasks—autobiographical events and public events—between those that he consciously remembered and those he somehow just knew about. Those that he consciously remembered represent the same conditions as those performed by the control subjects, with the nature of those he knew all about, but had no conscious recollection of, being unclear. When the autobiographical events he remembered were compared with autobiographical events known, the hippocampus bilaterally was more active for personal events he remembered experiencing: left hippocampus –24, 0, –27; 4.24; right hippocampus 15, 0, –29; 4.1 (Fig. 2A). The plot of the effect size from the peak voxel in the left hippocampus (Fig. 2B), where autobiographical event memory is represented by the first bar, clearly shows the differential response of the hippocampal region to this memory type. The location of this activation, although more anterior, is within the spatial extent of the hippocampal activation observed when autobiographical events were compared with autobiographical facts in the control subjects. The controls also activated the medial frontal region in this comparison, and indeed so did Jon (BA 11: 0, 57, –15; 5.02; and BA 10: –12, 54, 3; 4.9). This hippocampal/medial frontal pattern was seen when autobiographical events Jon remembers were compared with any other memory condition. Comparisons between the other memory types showed no significant differences in activity.
|
Regional interactions: effective connectivity
We have, so far, determined the anatomy of memory retrieval and the response patterns of specific regions in both control subjects and a patient with selective bilateral hippocampal damage. However, memory is not the property of brain regions operating in isolation, but rather of brain networks; thus, functional integration within this network must also be considered. Structural equation modelling was therefore applied to the fMRI data to estimate effective connectivity with the aim of assessing if connectivity between brain regions within the memory retrieval network changed as a function of the type of memory being recollected. This was examined in the control subjects and in patient Jon separately in the first instance, and then the two were directly compared. Group control data is reported here, although consistency in the findings was high across control subjects. Statistical inference was made using the
2 statistic to test the difference between the memory conditions/subject types (see Methods).
Analyses reported here focus on regional interactions within one hemisphere. Seven out of the total of nine regions activated in the main effect of memory compared with the control task in the control subjects were left-lateralized. The activity from the voxel of peak activation in each of these seven regions was included in the analysis of effective connectivity. Regions were: medial frontal cortex, temporal pole, hippocampus, anterior middle temporal gyrus, parahippocampal gyrus, retrosplenial cortex, temporoparietal junction. Although activations in Jon were significantly more bilateral than in control subjects, analysis of effective connectivity was also performed for the left hemisphere in Jon to afford direct comparison with the control subjects. Fig. 3A shows the model and connections which best characterized the memory retrieval data in the control subjects (see also Maguire et al., 2000).
|
Control subjects
Several significant changes in path coefficients emerged in the estimation of effective connectivity (Fig. 3B
). The first of these was the connection between parahippocampal cortex and hippocampus. When connectivity was compared for the two memory types that involve personally relevant memories, parahippocampal–hippocampal connectivity was found to increase significantly during the retrieval of autobiographical events (path coefficient = 0.5) relative to autobiographical facts (0.3) [2diff(1) = 4.9, P < 0.05]. Indeed, when comparison was made between autobiographical events and the other three memory types (0.3), parahippocampal–hippocampal connectivity was increased significantly during the retrieval of autobiographical events [2diff(1) = 4.1, P < 0.05]. Connectivity between parahippocampal cortex and temporal pole also increased during the retrieval of autobiographical events (0.3) compared with public events (–0.2) [2diff(1) = 7.8, P < 0.01]. In contrast, an increase in connectivity between lateral temporal cortex and temporal pole was found during the retrieval of public events (0.4) compared with autobiographical events (0.1) [2diff(1) = 3.6, P < 0.06]. The retrieval of general facts (0.3) relative to the retrieval of autobiographical events (0.1) also gave rise to an increase in connectivity between lateral temporal cortex and temporal pole [2diff(1) = 3.8, P < 0.05].
Patient Jon
Effective connectivity was then estimated for the left hemisphere of Jon. In the first instance, the model for the control subjects was applied to his data. However, extra connections were required in Jon to make the model stable (shown in blue on Fig. 3C), these being between retrosplenial–medial frontal cortices, retrosplenial cortex–hippocampus, and lateral temporal cortex–temporoparietal junction. For Jon, the only changes in the estimation of effective connectivity were in two connections during retrieval of autobiographical event memories he remembered compared with the other memory tasks (Fig. 3D), namely increases in connectivity between retrosplenial cortex and hippocampus (0.65 versus 0.02) [2diff(1) = 9.99, P < 0.002], and parahippocampal gyrus and temporal pole (0.48 versus 0.13) [2diff(1) = 6.65, P < 0.009]. When Jon's model was applied to the control subjects, the pattern of connectivity changes remained the same as with their original model.
Direct comparisons of the control subjects and Jon
Finally, the estimates of effective connectivity for Jon and the control subjects were directly compared, particularly for the three extra connections Jon required (Fig. 4). In the case of the retrosplenial cortex–hippocampal connection, effective connectivity was significantly increased for Jon (0.26) compared with the controls (–0.03) [2diff(1) = 3.86, P < 0.05] during retrieval of remembered autobiographical events in particular. This was also the case for the retrosplenial–medial frontal cortex connection (J: 0.32; C: –0.14) [2diff(1) = 6.65, P < 0.01]. The lateral temporal–temporoparietal junction connection was significantly increased in Jon (0.4) compared with the controls (0.03) during retrieval of general facts [2diff(1) = 3.9, P < 0.05].
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Using fMRI, we examined successful retrieval of real-world memories in a patient (Jon) with selective bilateral hippocampal pathology compared with healthy control subjects. Jon activated the same brain regions as control subjects, both medial and lateral on the left. In contrast to controls, Jon also activated many of the homologous regions on the right, and additionally several bilateral prefrontal areas. Despite 50% volume loss bilaterally in his hippocampi, memory retrieval in Jon was associated with increased activation of the hippocampi. Furthermore, activity in his hippocampi, as with the controls, was differential, being most responsive to retrieval of autobiographical events compared with the other memory types. The distinction Jon made between events that he clearly remembers and those he merely knows about appears to have a neural basis for autobiographical events. His hippocampi and medial frontal cortex were significantly more active during retrieval of events for which he had clear autonoetic conscious recollection compared with those he knew just as much about, including the context, but could not remember experiencing. Although Jon activates the same network of brain regions as the controls (albeit bilaterally), and with the same pattern of responses in the hippocampus, the communication between regions differs from controls with regard to hippocampal–cortical connectivity. The interplay between regions in response to autobiographical events in particular requires greater interaction in Jon between the hippocampus and retrosplenial cortex, and also increased extra-hippocampal interaction specifically between retrosplenial and medial frontal cortex.
Functional neuroimaging of patients who retain intact task performance provides rare insights not available from animal, structural neuroimaging, or neuropsychological data. In this case, we were able to identify commonalities and differences in the regions and patterns of activation between one such patient and healthy control subjects during real-world memory retrieval. When Jon successfully remembered something, all of the regions active in control subjects during memory retrieval were active in Jon (for discussion of specific regions comprising the network, see Maguire and Mummery, 1999; Maguire et al., 2000). Of particular interest here, the hippocampi of Jon, despite being damaged, were nevertheless functional (even though the tissue may not have been entirely normal; see T2 relaxometry findings reported in Vargha-Khadem et al., 1997). Indeed, given that Jon had less hippocampal volume to work with, so to speak, it may be that his hippocampi were more active, as there was no difference in activation when Jon and the controls were directly contrasted. This finding has important implications for the study of patients in the clinical context, demonstrating the possibility and importance of assessing the functionality of damaged brain tissue and its contribution to partially or completely retained functions. This has particular pertinence in the case of patients with selective hippocampal damage where the pervasive view is that significant damage renders such patients functionally ahippocampal (e.g. Rosenbaum et al., 2000). Our results show that this assumption should not automatically be made, as even with 50% volume, functional activation can occur in a manner similar to that seen in the healthy hippocampus.
We know that the hippocampal activations in Jon are functionally relevant, as activity was observed not just across all of the memory tasks, but also differentially, being most active during retrieval of autobiographical events, mirroring the pattern in control subjects. Findings from Jon enable us to refine further the contribution of the hippocampus in memory retrieval beyond that discernible from control subjects. In controls, autobiographical events were always accompanied by conscious recollection of the episodes. In Jon, he made a distinction between those events that he clearly remembered and those that he merely knew about, even though they were matched in terms of amount and type of details recalled. Jon's data show that it is not enough for a memory to be an autobiographical event about which one can recall rich details; there must also be recollection of the experience in order to be of particular interest to the hippocampus, in line with the view of Tulving and Markowitsch (1998), Aggleton and Brown (1999) and Eldridge et al. (2000). In the case of real-world autobiographical event memories, therefore, it would seem that there is a neural basis to the remember/know distinction. It is notable that in Jon the same distinction for public events revealed no differences and no specific activation of the hippocampus. It is possible that inconsistent remember/know retrieval findings in the literature are due to differences in the type, complexity or realism of stimuli, and the extent to which subjects process them as autobiographical events. A related issue for future research concerns whether a similar functional anatomical distinction could be observed for remember/know memories in control subjects, where amount and type of details about events are matched, as, for example, in the case of real versus false memories (Dodson et al., 2000).
In control subjects and Jon, autobiographical event memory compared with the other memory types was associated with medial frontal as well as hippocampal activation. Frontal regions have been implicated in autobiographical memory; for example, Hodges and Gurd (1994) report a deficit in self-relevant memory in a patient with frontal lobe Pick's disease. Tasks requiring consignment of `agency' to others have activated a medial frontal region similar to that noted here (Frith, 1996; Gallagher et al., 2000). In the present study, this region may have been involved in the verification of memory in reference to oneself, consonant with its proposed role in the memory supervisory system (Shallice, 1988). Jon activated several other prefrontal regions not apparent in control subjects, namely superior and inferior frontal gyri on the right, and middle frontal gyrus on the left. These activations cannot be attributed to differences between Jon and the control subjects in terms of retrieval success (Rugg et al., 1996), as this was comparable, but may reflect more effort (see Rugg and Wilding, 2000) on the part of the patient and/or greater use of working memory to maintain online the task instructions (Rowe et al., 2000).
Another difference between Jon and the control subjects is the bilateral nature of Jon's activations. This is unlikely to be solely due to differences in attention, retrieval effort or working memory, as the areas on the right are homologous to those activated in control subjects, many of them not primarily associated with such functions. Neither is it the case that the data reflect switching (as they are bilateral) or recruitment of areas immediately adjacent to damaged regions. The reason for the bilateral representation is not discernible from the present data. We can suggest that the recruitment of the right hemisphere may reflect additional use of visual strategies in Jon compared with the controls. It might also be a reflection of the early nature of Jon's damage, where, in order to compensate for impoverished learning, both hemispheres were primed from the outset to process signals and boost performance. One way to examine whether his bilateral pattern is due to plastic changes early on would be to scan a patient with similarly selective anoxic damage to the hippocampi acquired in adulthood to see if a more unilateral control-like picture is apparent, and we are currently exploring this issue.
The final difference to emerge between Jon and the control subjects was apparent at a level of analysis less commonly tapped in neuroimaging studies, namely regional interactions. By estimating effective connectivity between the brain regions active during memory retrieval, we were able to examine changes in connectivity due to memory type and also due to subject status (i.e. control subject or patient). An illustration of the importance of separating measurement of changes in regional brain activation from changes in inter-regional communication is evident in the report by Büchel and colleagues (Büchel et al., 1999). They found that, although activation in a cortical area decreased over time in association with the repeated performance of a task, the strength of functional connections actually increased between this and other brain regions. Thus, much importance may lie in the functional connections between brain regions as well as the changes in activity within regions. We observed increased connectivity associated with retrieval of autobiographical events compared with other memory types in the control subjects between parahippocampal gyrus and hippocampus, and parahippocampal gyrus and temporal pole (Maguire et al., 2000).
In the first instance, extra connections were required in Jon's model compared with the controls. This reflected more inputs from retrosplenial cortex into both medial temporal and medial frontal areas. It is known that the retrosplenial cortex is involved in memory, as it is activated in many neuroimaging memory studies (e.g. Maddock, 1999; Maguire et al., 1999; see also Maguire, 2001), and cases of amnesia following lesions involving this area have been reported (Valenstein et al., 1987). For Jon, the only changes in the estimation of effective connectivity were in two connections during retrieval of autobiographical events he remembered compared with the other memory types, namely increased connectivity between retrosplenial cortex and hippocampus, and between parahippocampal gyrus and temporal pole. Although he has the latter change in connectivity in common with the control subjects, there was no change in connectivity between the parahippocampal gyrus and hippocampus in Jon as observed in control subjects during autobiographical event retrieval. As well as activity within medial temporal lobe structures, the interplay between them, particularly the hippocampus and parahippocampal cortex, may be important for supporting the recollection of autobiographical episodes. Thus, although Jon's regional activations mirror the controls, and his hippocampi are at least partially functional, the absence of increased hippocampal–parahippocampal connectivity during retrieval of autobiographical events may be an important clue to the nature of his memory disorder. Perhaps the damage to his hippocampi has an impact on the connectivity with parahippocampal cortex—a major communication route to and from the hippocampus. Instead there is in Jon increased retrosplenial–hippocampal connectivity. That this is possibly not an optimal pathway and may be capacity limited may explain Jon's ability to remember only a limited number of autobiographical events in the context of a general deficit for such memories. [As indicated earlier (see Introduction), Jon had only very few autobiographical event memories, but enough to make this study viable.] One further point should be noted in relation to the known anatomical connectivity of the retrosplenial cortex. It has connections with both parahippocampal and entorhinal cortices as well as presubiculum (for details, see Morris et al., 1999; Kobayashi and Amaral, 2000). If access to the hippocampus via the parahippocampal gyrus is diminished (as suggested by the lack of increased connectivity between them in Jon as compared with the controls), then perhaps the retrosplenial cortex is interacting in Jon with presubiculum. Given the current spatial resolution, this cannot be ruled out and remains to be probed further.
It is also notable that when compared with control subjects, the retrosplenial–medial frontal connectivity increased for Jon during autobiographical event memories. Since medial frontal cortex as well as the hippocampus was activated in both Jon and the controls during autobiographical event retrieval in particular, it may be that the increased connectivity with this cortex in Jon reflects the use of another relevant but suboptimal route for signals that would normally go via the hippocampus.
There are caveats associated with the estimation of effective connectivity. For example, the models are simple and it is difficult to infer whether changes reflect excitatory or inhibitory influences, and detailed knowledge of human neuroanatomical connectivity is still lacking. Notwithstanding current constraints, the combination of modelling the real time-course of brain activity while subjects use a memory retrieval network to recollect memories peculiar to themselves adds another dimension to our study of memory. This is particularly pertinent in the context of memory breakdown where some impairments may be detectable only at the level of inter-regional communication, as in the case of Jon, or the earliest stages of disease (e.g. dementia).
Whilst we were able to capitalize on the fact that Jon does retain some autobiographical event memories, the majority have clearly not been retained. An obvious question is why some events get remembered and others do not, or why some are known, but the experience not remembered. The present study does not speak to the issue, for example, of whether it is due to problems at initial encoding, consolidation/storage or retrieval. Here, the focus was on retrieval. It may be that the events Jon remembers were rehearsed more at encoding and so were more deeply processed (Craik and Lockhart, 1972), or rehearsed more since occurring, although we specifically selected events that his mother indicated were not repeatedly discussed subsequent to their initial occurrence. Or perhaps some events were more emotionally salient, although we made specific attempts to match the events for this variable. After scanning, we also asked Jon to rate the emotional valence and intensity evoked by recalling the events that were used for the scanning experiment (in a similar manner to Lane et al., 1999). The valence (using a scale of –3 to +3) did not differ significantly (at an explicit level) between the events that were known (1.13) and remembered (1.56), nor did intensity (on a scale of 1–7) differ between those known (4.19) and remembered (4.19). Understanding Jon's anterograde episodic memory deficit and how it interacts with his retrieval deficits is clearly an important issue for future research.
In conclusion, Jon's case makes clear the value of scanning patients using fMRI during tasks they can perform. Applying this method allowed us to confirm the functionality of remnant tissue in a damaged structure, and to appreciate the neural basis of a cognitive distinction (remember/know) that control subjects did not make, and, in doing so to refine further knowledge of the hippocampal role in autobiographical event memory. While this is a single case, and clearly the results need to be replicated in other similar patients, this study also indicates that plasticity at the level of recruitment of bilateral brain regions and altered patterns of effective connectivity between brain regions may be important indicators of disordered memory. Whether this plasticity reflects the reorganizational capacity of the immature brain or the selectivity of the hippocampal pathology has yet to be determined.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
We are indebted to Jon and his parents for their cooperation in this study. We wish to thank C. J. Mummery for assistance with control data collection, and C. Büchel for advice on effective connectivity. We also wish to thank C. Frith and R. Frackowiak for helpful comments. E.A.M. is supported by the Wellcome Trust. This work was undertaken in part by Great Ormond Street Hospital for Children NHS Trust, which received a proportion of its funding from the NHS Executive; the views expressed in this publication are those of the authors and are not necessarily those of the NHS Executive.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Aggleton JP, Brown MW. Episodic memory, amnesia, and the hippocampal-anterior thalamic axis. Behav Brain Sci 1999; 22: 425–89.[Web of Science][Medline]
Amaral DG, Insausti R. Hippocampal formation. In: Paxinos G, editor. The human nervous system. San Diego: Academic Press; 1990. p. 711–55.
Arnold SE, Hyman BT, Van Hoesen GW. Neuropathologic changes of the temporal pole in Alzheimer's disease and Pick's disease. Arch Neurol 1994; 51: 145–50.
Bollen KA. Structural equations with latent variables. New York: Wiley; 1989.
Büchel C, Friston KJ. Modulation of connectivity in visual pathways by attention: cortical interactions evaluated with structural equation modelling and fMRI. Cereb Cortex 1997; 7: 768–78.
Büchel C, Coull JT, Friston KJ. The predictive value of changes in effective connectivity for human learning. Science 1999; 283: 1538–41.
Conway MA, Turk DJ, Miller SL, Logan J, Nebes RD, Meltzer CC, et al. A positron emission tomography (PET) study of autobiographical memory retrieval. Memory 1999; 7: 679–702.[Web of Science][Medline]
Craik FIM, Lockhart RS. Levels of processing: a framework for memory research. J Verb Learn Verb Behav 1972; 11: 671–84.
Dodson CS, Koutstaal W, Schacter DL. Escape from illusion: reducing false memories. Trends Cogn Sci 2000; 4: 391–7.[Web of Science][Medline]
Eldridge LL, Knowlton BJ, Furmanski CS, Bookheimer SY, Engel SA. Remembering episodes: a selective role for the hippocampus during retrieval. Nat Neurosci 2000; 3: 1149–52.[Web of Science][Medline]
Evans AC, Collins DL, Mills SR, Brown ED, Kelly RL, Peters TM. 3D statistical neuroanatomical models from 305 MRI volumes. In: Klaisner LA, editor. Proceedings of the IEEE-Nuclear Science Symposium and Medical Imaging Conference. Piscataway (NJ): IEEE Service Center; 1993. p. 1813–17.
Fink GR, Markowitsch HJ, Reinkemeier M, Bruckbauer T, Kessler J, Heiss WD. Cerebral representation of one's own past: neural networks involved in autobiographical memory. J Neurosci 1996; 16: 4275–82.
Fletcher P, Büchel C, Josephs O, Friston K, Dolan R. Learning-related neuronal responses in prefrontal cortex studied with functional neuroimaging. Cereb Cortex 1999; 9: 168–78.
Friston KJ, Frith CD, Frackowiak RSJ. Time-dependent changes in effective connectivity measured with PET. Hum Brain Mapp 1993; 1: 69–79.
Friston KJ, Holmes AP, Worsley KJ, Poline JB, Frith CD, Frackowiak RSJ. Statistical parametric maps in functional imaging: a general linear approach. Hum Brain Mapp 1995; 2: 189–210.
Frith C. The role of the prefrontal cortex in self-consciousness: the case of auditory hallucinations. [Review]. Philos Trans R Soc London B Biol Sci 1996; 351: 1505–12.[Web of Science][Medline]
Gadian DG, Aicardi J, Watkins KE, Porter DA, Mishkin M, Vargha-Khadem F. Developmental amnesia associated with early hypoxic-ischaemic injury. Brain 2000; 123: 499–507.
Gallagher HL, Happe F, Brunswick N, Fletcher PC, Frith U, Frith CD. Reading the mind in cartoons and stories: an fMRI study of `theory of mind' in verbal and nonverbal tasks. Neuropsychologia 2000; 38: 11–21.[Web of Science][Medline]
Gardiner JM, Java RI. Forgetting in recognition memory with and without recollective experience. Mem Cognit 1991; 19: 617–23.[Web of Science][Medline]
Henson RN, Rugg MD, Shallice T, Josephs O, Dolan RJ. Recollection and familiarity in recognition memory: an event-related functional magnetic resonance imaging study. J Neurosci 1999; 19: 3962–72.
Hirano M, Noguchi K. Dissociation between specific personal episodes and other aspects of remote memory in a patient with hippocampal amnesia. Percept Mot Skills 1998; 87: 99–107.[Web of Science][Medline]
Hodges JR, Gurd JM. Remote memory and lexical retrieval in a case of frontal Pick's disease. Arch Neurol 1994; 51: 821–7.
Holmes AP, Josephs O, Büchel C, Friston KJ. Statistical modelling of low-frequency confounds in fMRI. Neuroimage 1997; 5 (4 Pt 2): S480.
Horwitz B, Tagamets M-A, McIntosh AR. Neural modeling, functional brain imaging, and cognition. Trends Cogn Sci 1999; 3: 91–8.[Web of Science][Medline]
Kapur N, Pearson D. Memory symptoms and memory performance of neurological patients. Br J Psychol 1983; 74: 409–15.
Kitchener EG, Hodges JR, McCarthy R. Acquisition of post-morbid vocabulary and semantic facts in the absence of episodic memory. Brain 1998; 121: 1313–27.
Knowlton BJ, Squire LR. Remembering and knowing: two different expressions of declarative memory. J Exp Psychol Learn Mem Cogn 1995; 21: 699–710.[Web of Science][Medline]
Kobayashi Y, Amaral DG. Macaque monkey retrosplenial cortex: I. Three-dimensional and cytoarchitectonic organization. J Comp Neurol 2000; 426: 339–65.[Web of Science][Medline]
Köhler S, McIntosh AR, Moscovitch M, Winocur G. Functional interactions between the medial temporal lobe and posterior neocortex related to episodic memory retrieval. Cereb Cortex 1998; 8: 451–61.
Lane RD, Chua PM, Dolan RJ. Common effects of emotional valence, arousal and attention on neural activation during visual processing of pictures. Neuropsychologia 1999; 37: 989–97.[Web of Science][Medline]
Levine B, Black SE, Cabeza R, Sinden M, McIntosh AR, Toth JP, et al. Episodic memory and the self in a case of isolated retrograde amnesia. Brain 1998; 121: 1951–73.
Maddock RJ. The retrosplenial cortex and emotion: new insights from functional neuroimaging of the human brain. [Review]. Trends Neurosci 1999; 22: 310–6.[Web of Science][Medline]
Maguire EA. The retrosplenial contribution to human navigation: a review of lesion and neuroimaging findings. Scand J Psychol. In press 2001.
Maguire EA, Mummery CJ. Differential modulation of a common memory retrieval network revealed by positron emission tomography. Hippocampus 1999; 9: 54–61.[Web of Science][Medline]
Maguire EA, Frith CD, Morris RG. The functional neuroanatomy of comprehension and memory: the importance of prior knowledge. Brain 1999; 122: 1839–50.
Maguire EA, Mummery CJ, Büchel C. Patterns of hippocampal-cortical interaction dissociate temporal lobe memory subsystems. Hippocampus 2000; 10: 475–82.[Web of Science][Medline]
Maguire EA, Henson RNA, Mummery CJ, Frith CD. Activity in prefrontal cortex, not hippocampus, varies parametrically with the increasing remoteness of memories. Neuroreport 2001; 12: 441–4.[Web of Science][Medline]
Manns JR, Squire LR. Impaired recognition memory on the Doors and People Test after damage limited to the hippocampal region. Hippocampus 1999; 9: 495–9.[Web of Science][Medline]
McCarthy RA, Warrington EK. Actors but not scripts: the dissociation of people and events in retrograde amnesia. Neuropsychologia 1992; 30: 633–44.[Web of Science][Medline]
McIntosh AR, Cabeza RE, Lobaugh NL. Analysis of neural interactions explains the activation of occipital cortex by an auditory stimulus. J Neurophysiol 1998; 80: 2790–6.
Morris R, Pandya DN, Petrides, M. Fiber system linking the mid-dorsolateral frontal cortex with the retrosplenial/presubicular region in the rhesus monkey. J Comp Neurol 1999; 407: 183–92.[Web of Science][Medline]
Nyberg L, McIntosh AR, Cabeza R, Habib R, Houle S,Tulving E. General and specific brain regions involved in encoding and retrieval of events: what, where, and when. Proc Natl Acad Sci USA 1996; 93: 11280–5.
O'Keefe J, Nadel L. The hippocampus as a cognitive map. Oxford: Clarendon Press; 1978.
Price CJ, Friston KJ. Scanning patients with tasks they can perform. Hum Brain Mapp 1999; 8: 102–8.[Web of Science][Medline]
Reed LJ, Marsden P, Lasserson D, Sheldon N, Lewis P, Stanhope N, et al. FDG-PET analysis and findings in amnesia resulting from hypoxia. Memory 1999; 7: 599–612.[Web of Science][Medline]
Rosenbaum RS, Priselac S, Köhler S, Black SE, Gao F, Nadel L, et al. Remote spatial memory in an amnesic person with extensive bilateral hippocampal lesions. Nat Neurosci 2000; 3: 1044–8.[Web of Science][Medline]
Rowe JB, Toni I, Josephs O, Frackowiak RS, Passingham RE. The prefrontal cortex: response selection or maintenance within working memory? Science 2000; 288: 1656–60.
Rugg MD, Wilding EL. Retrieval processing and episodic memory. Trends Cogn Sci 2000; 4: 108–15.[Web of Science][Medline]
Rugg MD, Fletcher PC, Frith CD, Frackowiak RS, Dolan RJ. Differential activation of the prefrontal cortex in successful and unsuccessful memory retrieval. Brain 1996; 119: 2073–83.
Rugg MD, Fletcher PC, Chua PM, Dolan RJ. The role of the prefrontal cortex in recognition memory and memory for source: an fMRI study. Neuroimage 1999; 10: 520–9.[Web of Science][Medline]
Scoville WB, Milner B. Loss of recent memory after bilateral hippocampal lesions. J Neurol Neurosurg Psychiatry 1957; 20: 11–21.
Shallice T. From neuropsychology to mental structure. Cambridge: Cambridge University Press; 1988.
Squire LR. Memory and the hippocampus: a synthesis from findings with rats, monkeys, and humans. [Review]. Psychol Rev 1992; 99: 195–231.[Web of Science][Medline]
Swartz BE, Mandelkern MA. Positron emission tomography: the contribution of cognitive activation paradigms to the understanding of the epilepsies. [Review]. Adv Neurol 1999; 79: 901–15.[Medline]
Talairach J, Tournoux P. Co-planar stereotaxic atlas of the human brain. Stuttgart: Thieme; 1988.
Tulving E. Elements of episodic memory. Oxford: Clarendon Press; 1983.
Tulving E. Memory and consciousness. Can Psychol 1985; 26: 1–12.
Tulving E, Markowitsch HJ. Episodic and declarative memory: role of the hippocampus. [Review]. Hippocampus 1998; 8: 198–204.[Web of Science][Medline]
Valenstein E, Bowers D, Verfaellie M, Heilman KM, Day A, Watson RT. Retrosplenial amnesia. Brain 1987; 110: 1631–46.
Van Hoesen GW. The parahippocampal gyrus: new observations regarding its cortical connections in the monkey. Trends Neurosci 1982; 5: 345–50.[Web of Science]
Van Hoesen GW. Ventromedial temporal lobe anatomy, with comments on Alzheimer's disease and temporal injury. [Review]. J Neuropsychiatry Clin Neurosci 1997; 9: 331–41.
Vargha-Khadem F, Gadian DG, Watkins KE, Connelly A, Van Paesschen W, Mishkin M. Differential effects of early hippocampal pathology on episodic and semantic memory. Science 1997; 277: 376–80.
Received December 15, 2000. Revised February 6, 2001. Accepted February 9, 2001.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. Dierssen, Y. Herault, and X. Estivill Aneuploidy: From a Physiological Mechanism of Variance to Down Syndrome Physiol Rev, July 1, 2009; 89(3): 887 - 920. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. L. Harenski, O. Antonenko, M. S. Shane, and K. A. Kiehl Gender differences in neural mechanisms underlying moral sensitivity Soc Cogn Affect Neurosci, December 1, 2008; 3(4): 313 - 321. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. A. Mittal, L. M. Ellman, and T. D. Cannon Gene-Environment Interaction and Covariation in Schizophrenia: The Role of Obstetric Complications Schizophr Bull, November 1, 2008; 34(6): 1083 - 1094. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Braun, C. Finke, F. Ostendorf, T.-N. Lehmann, K.-T. Hoffmann, and C. J. Ploner Reorganization of associative memory in humans with long-standing hippocampal damage Brain, October 1, 2008; 131(10): 2742 - 2750. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. W. Davis, N. A. Dennis, S. M. Daselaar, M. S. Fleck, and R. Cabeza Que PASA? The Posterior-Anterior Shift in Aging Cereb Cortex, May 1, 2008; 18(5): 1201 - 1209. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. R. Addis, M. Moscovitch, and M. P. McAndrews Consequences of hippocampal damage across the autobiographical memory network in left temporal lobe epilepsy Brain, September 1, 2007; 130(9): 2327 - 2342. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Hassabis, D. Kumaran, S. D. Vann, and E. A. Maguire Patients with hippocampal amnesia cannot imagine new experiences PNAS, January 30, 2007; 104(5): 1726 - 1731. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.-P. Vignal, L. Maillard, A. McGonigal, and P. Chauvel The dreamy state: hallucinations of autobiographic memory evoked by temporal lobe stimulations and seizures Brain, January 1, 2007; 130(1): 88 - 99. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. A. Maguire, R. Nannery, and H. J. Spiers Navigation around London by a taxi driver with bilateral hippocampal lesions. Brain, November 1, 2006; 129(Pt 11): 2894 - 2907. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B Guillery-Girard, B Desgranges, C Urban, P Piolino, V de la Sayette, and F Eustache The dynamic time course of memory recovery in transient global amnesia J. Neurol. Neurosurg. Psychiatry, November 1, 2004; 75(11): 1532 - 1540. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Gilboa, G. Winocur, C. L. Grady, S. J. Hevenor, and M. Moscovitch Remembering Our Past: Functional Neuroanatomy of Recollection of Recent and Very Remote Personal Events Cereb Cortex, November 1, 2004; 14(11): 1214 - 1225. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Lustig, A. Z. Snyder, M. Bhakta, K. C. O'Brien, M. McAvoy, M. E. Raichle, J. C. Morris, and R. L. Buckner Functional deactivations: Change with age and dementia of the Alzheimer type PNAS, November 25, 2003; 100(24): 14504 - 14509. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Vargha-Khadem, C. H. Salmond, K. E. Watkins, K. J. Friston, D. G. Gadian, and M. Mishkin Developmental amnesia: Effect of age at injury PNAS, August 19, 2003; 100(17): 10055 - 10060. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. R. Fink In search of one's own past: the neural bases of autobiographical memories Brain, July 1, 2003; 126(7): 1509 - 1510. [Full Text] [PDF]
|
|
|
|
|
|
E. A. Maguire and C. D. Frith Aging affects the engagement of the hippocampus during autobiographical memory retrieval Brain, July 1, 2003; 126(7): 1511 - 1523. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. A. Maguire and C. D. Frith Lateral Asymmetry in the Hippocampal Response to the Remoteness of Autobiographical Memories J. Neurosci., June 15, 2003; 23(12): 5302 - 5307. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Kapur and M. D Kopelman Advanced brain imaging procedures and human memory disorder Br. Med. Bull., March 1, 2003; 65(1): 61 - 81. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Piefke, P. H. Weiss, K. Zilles, H. J. Markowitsch, and G. R. Fink Differential remoteness and emotional tone modulate the neural correlates of autobiographical memory Brain, March 1, 2003; 126(3): 650 - 668. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. C. Johnson, L. C. Baxter, L. S. Wilder, J. G. Pipe, J. E. Heiserman, and G. P. Prigatano Neural correlates of self-reflection Brain, August 1, 2002; 125(8): 1808 - 1814. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. J. Curtis, L. L. Lindeke, M. K. Georgieff, and C. A. Nelson Neurobehavioural functioning in neonatal intensive care unit graduates in late childhood and early adolescence Brain, July 1, 2002; 125(7): 1646 - 1659. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||