Abstract

Using event-related fMRI, we scanned young healthy subjects while they memorized real-world photographs and subsequently tried to recognize them within a series of new photographs. We confirmed that activity in the medial temporal lobe (MTL) and inferior prefrontal cortex correlates with declarative memory formation as defined by the subsequent memory effect, stronger responses to subsequently remembered than forgotten items. Additionally, we confirmed that activity in specific regions within the parietal lobe, anterior prefrontal cortex, anterior cingulate and cerebellum correlate with recognition memory as measured by the conventional old/new effect, stronger responses for recognized old items (hits) than correctly identified new items (correct rejections). To obtain a purer measure of recognition success, we introduced two recognition effects by comparing brain responses to hits and old items misclassified as new (misses). The positive recognition effect (hits > misses) revealed prefrontal, parietal and cerebellar contributions to recognition, and in line with electrophysiological findings, the negative recognition effect (hits < misses) revealed an anterior medial temporal contribution. Finally, by inclusive masking, we identified temporal and cerebellar brain areas that support both declarative memory formation and retrieval. For matching operations during recognition, these areas may re-use representations formed and stored locally during encoding.

Introduction

The kind of memory one ordinarily means when using the term ‘memory’ is declarative memory, which enables us to consciously remember past events and facts (Cohen and Squire, 1980). Declarative memory is based on at least two fundamental mnemonic operations: memory formation and retrieval (Gabrieli, 1998). Since only a few years ago, event-related functional magnetic resonance imaging (ER-fMRI) has provided the unique opportunity to study the neural correlates of these mnemonic operations with great anatomical detail in healthy human subjects (Dale and Buckner, 1997; Josephs et al., 1997; Zarahn et al., 1997). Using ER-fMRI, encoding studies have shown that successful declarative memory formation, measured as the difference in brain activity between subsequently remembered and forgotten items, is accompanied by activity increases in medial temporal and inferior prefrontal areas (e.g. Brewer et al., 1998; Wagner et al., 1998; Kirchhoff et al., 2000; Davachi et al., 2001; Otten et al., 2001; Otten and Rugg, 2001a; Strange et al., 2002; Morcom et al., 2003: for a review, see Paller and Wagner, 2002) and activity decreases in posterior cingulate, parietal and dorsolateral prefrontal areas (Otten and Rugg, 2001b). ER-fMRI studies acquiring fMRI data during simple recognition memory tests and applying the so-called old/new effect, the difference in brain activity between correctly recognized old, previously studied items (hits) and correctly identified new, previously unstudied items (correct rejections), have shown activations in the anterior prefrontal cortex, parietal cortex, insula and medial-frontal areas including the anterior cingulate (e.g. Henson et al., 1999; Konishi et al., 2000; Donaldson et al., 2001a,b; for a review, see Rugg and Henson, 2002). The first aim of the present study is to replicate these subsequent memory and old/new effects, which were so far obtained in separate encoding and retrieval experiments, within a single study-test experiment.

Based on this empirical foundation, we aim to explore in the second step of this study whether recognition success can be associated with both regional brain activity increases and decreases. Brain activity increases for hits as compared to correct rejections (old/new effect) have been interpreted as related to the successful recovery of information from declarative memory (Donaldson and Buckner, 1999; Konishi et al., 2000; Donaldson et al., 2001a,b). A reversed old/new contrast, however, cannot delineate cleanly a brain activity decrease related to recognition success, because it would be heavily contaminated by neural correlates of repetition priming (Buckner and Koutstaal, 1998; Donaldson et al., 2001a). Repetition priming is an implicit memory phenomenon that improves processing efficacy of repeatedly processed items and that is regularly accompanied by weaker brain activity to old as compared to new items (Tulving and Schacter, 1990; but see Henson et al., 2000). However, repetition priming does not support conscious recognition (Donaldson et al., 2001a). Thus, the question that remains open is: can only repetition priming or also conscious recognition correlate with a decrease in neural activity? (Henson et al., 2003). Electrophysiological data in animals and humans suggest that recognition can also be accompanied by brain activity decreases (Smith et al., 1986; Brown et al., 1987; Miller and Desimone, 1994; Brown and Aggleton, 2001; Fernández et al., 2001). As outlined above, we cannot simply reverse the old/new effect. Rather, analogous to the subsequent memory effect, we compare brain activity to correctly recognized old items (hits) and old items misclassified as new (misses). In this contrast, henceforth called the recognition effect, all items are studied once before, but recognition success differs. Hence, a negative recognition effect (misses > hits) seems to be less contaminated by repetition priming than a reversed old/new effect, at least when primed and recognized items show stochastic independence in the sense that performance in the two tasks is uncorrelated at the level of individual items (Shimamura, 1985). And in addition, the recognition effect might generally be more closely related to recognition success than the old/new effect, because it does not include any difference related to the actual study status of the items. Thus, as our second goal, we aim to identify increases and decreases in brain activity associated with recognition success as indexed by a positive (hits > misses) and a negative recognition effect (misses > hits). A recent meta-analysis of four event-related fMRI studies employing different kinds of study material suggested that less anterior MTL activity is related to the amount of familiarity across a variety of stimulus materials (Henson et al., 2003). Another event-related fMRI study (Rombouts et al., 2001) found anterior parahippocampal gyrus activation in a comparison of new to often seen items, but this study did not control for performance. In line with these findings as well as with electrophysiological data (Smith et al., 1986; Miller and Desimone, 1994; Brown and Aggleton, 2001; Fernández et al., 2001), we expect negative recognition effects in inferior temporal areas including the anterior MTL.

Given the encoding and recognition results of ER-fMRI studies reported above, there seems to be no or almost no overlap between brain areas involved in both memory formation and recognition (see also Gabrieli et al., 1997). If, however, a brain area would support these two operations, neural representations stored locally during encoding could be re-used during recognition. Such a module would not only be efficient and intuitive, its existence is supported by electrophysiological data. For instance, the so called anterior MTL-N400, a negative component in event related potentials recorded invasively in epilepsy patients from the anterior MTL, probably from the perirhinal cortex (McCarthy et al., 1995) shows an amplitude difference between subsequently remembered and forgotten items during encoding (Fernández et al., 1999, 2002) as well as between correctly identified old and new items during a recognition memory test (Smith et al., 1986). Therefore, this neural node within the anterior MTL seems to be critically involved in both memory formation and retrieval. However, most studies to date have examined either memory encoding or retrieval and have therefore not been able directly to compare encoding- and retrieval-related activations within subjects. The third aim of the present fMRI study is thus to characterize this node within a single study-test experiment by a functional imaging approach applying inclusively masking of the subsequent memory effect and either the positive or the negative recognition effect. Moreover, we aim to identify further brain areas whose activity is correlated with both successful memory formation and recognition by whole brain coverage.

Material and Methods

Subjects

Sixteen healthy volunteers (eight male, eight female) with normal or corrected-to-normal vision participated in the experiment. All subjects were consistent right-handers according to the Edinburgh Handedness Index (mean EHI = 88, range 73–100; Oldfield, 1971). Their mean age was 30 years with a range of 20–49 years. Following approval by the Medical Ethics Committee of the University of Bonn, all subjects gave their written informed consent according to the Declaration of Helsinki (1991). They were paid for their participation.

Stimuli and Task

Stimuli consisted of 480 color photographs of either buildings or natural landscapes without any buildings (240 for each category) that were selected to be similar in complexity, brightness and contrast.

In the study phase, 120 randomly selected pictures of buildings were randomly intermixed with 120 pictures of landscapes. Stimuli were presented sequentially for 800 ms each with a randomized interstimulus interval (ISI) of 2000–3000 ms (mean 2500 ms). Sixty null events, consisting of a black screen shown for 2500 ms, were randomly intermixed. Subjects were required to memorize each picture and to make a building–landscape decision by right-hand key-press.

In the following recognition phase, all stimuli from the study phase plus 240 new, previously not presented photographs of buildings and landscapes were shown sequentially and again randomly intermixed. The presentation rate was self-paced by subjects’ responses, resulting in a mean ISI of 2065 ms (SD 229 ms). Subjects were required to press one of three keys according to the following response categories: picture seen before with high confidence, picture uncertain to be seen before or not, picture not seen before with high confidence.

Stimuli were presented using the Experimental Run-time System (http://www.erts.de) and back-projected onto a translucent screen positioned opposite the magnet bore using an LCD-projector. Subjects viewed the stimuli by way of a mirror mounted on the head coil while lying in a supine position with their head stabilized by an individually molded vacuum cushion.

fMRI Data Acquisition

All scans were performed on a 1.5 T scanner (Symphony; Siemens, Erlangen, Germany) using standard gradients and a circular polarized phase array head coil. For each subject, we acquired two series, one for the study phase and one for the recognition phase, of T2*-weighted axial EPI-scans including eight initial dummy scans parallel to the AC/PC line with the following parameters: number of slices (NS), 30; slice thickness (ST), 4 mm; interslice gap (IG), 0.4 mm; matrix size (MS), 64 × 64; field of view (FOV), 220 mm; echo time (TE), 50 ms; repetition time (TR), 2.95 s. The encoding run comprised 282 scans per subject. During recognition, we acquired 282–402 scans per subject (mean number of scans: 336), depending on the individual response times. T1-weighted 3D-FLASH scans were acquired between the functional runs for anatomical localization (NS = 120; ST = 1.5 mm; IG = none; MS = 256 × 256; FOV = 230 mm; TE = 4 ms; TR = 11 ms).

fMRI Data Analysis

MR images were analyzed using Statistical Parametric Mapping (SPM99; www.fil.ion.ucl.ac.uk) implemented in MATLAB (Mathworks Inc., Sherborn, MA). To correct for their different acquisition times, the signal measured in each slice was shifted relative to the acquisition time of the middle slice using a sinc interpolation in time. All images were realigned to the first image to correct for head movement and normalized into standard stereotaxic anatomical MNI-space by using the transformation matrix calculated from the first EPI-scan of each subject and the EPI-template. Afterwards, the normalized data with a resliced voxel size of 4 × 4 × 4 mm were smoothed with a 8 mm FWHM isotropic Gaussian kernel to accommodate intersubject variation in brain anatomy. Proportional scaling with high pass filtering was used to eliminate confounding effects of differences in global activity within and between subjects. All analyses were restricted to trials on which encoding responses were correct. The expected hemodynamic response at stimulus onset for each event-type was modeled by two response functions, a canonical hemodynamic response function (HRF; Friston et al., 1998) and its temporal derivative. The temporal derivative was included in the model to account for the residual variance resulting from small temporal differences in the onset of the hemodynamic response, which is not explained by the canonical HRF alone. The functions were convolved with the event-train of stimulus onsets to create covariates in a general linear model. During recognition, the presentation rate was self-paced, hence reaction time to recognition trials was included in the model as a nuisance variable to discount the possibility of a confounding effect of differences in reaction times. Parameter estimates for the HRF regressor were calculated from the least mean squares fit of the model to the time series. Parameter estimates for the temporal derivative were not considered in any contrast. Subsequently, effects of interest were specified by appropriately weighted linear contrasts of the HRF parameter estimates and determined using planned comparisons on a voxel-by-voxel basis; the corresponding linear combination of parameter estimates for each contrast were stored as separate images for each subject. For the sake of our study goals, trials related to ‘uncertain’ responses were modeled by a separate regressor, but not considered in any contrast.

An SPM99 group analysis was performed by entering contrast images into one-sample t-tests, in which subjects are treated as random variables. Voxels with a significance level of P < 0.005 uncorrected belonging to clusters with at least 10 voxels are reported. Since the subsequent memory effect and the recognition effect were computed bi-directionally, this effectively results in a two-sided threshold of P < 0.01. Activations are shown projected onto selected coronal slices of the mean high-resolution T1-weighted volume, highlighting regions of interest. The reported voxel coordinates of activation peaks were transformed from MNI space to Talairach and Tournoux (1988) atlas space by non-linear transformations (http://www.mrc-cbu.cam.ac.uk/imaging/mnispace.html). To address further the question of overlap between areas involved in both encoding and recognition, we recomputed the subsequent memory effect and the recognition effects with a statistical threshold of P < 0.05. Then we inclusively masked the subsequent memory effect by the positive recognition effect and the negative recognition effect respectively. These analyses permit the identification, at a high level of sensitivity, of regions in which the subsequent memory effect overlaps with recognition effects, while maintaining an acceptable type I error rate. According to Fisher’s method of combining probabilities, the probability of two independent statistical tests conjointly attaining significance at P < 0.05 is P < 0.017.

Results

Behavioral Results

During encoding, the building–landscape decision task was made with a mean accuracy of 92% (range 85–98%). Incorrect responses were recorded for 5% (2–10%) and no responses for 3% (0–7%) of all encoding trials.

Recognition memory performance and reaction times are listed in Table 1. Accuracy of recognition was assessed by the difference in probabilities of a correct old judgment and an old judgment for a new item (Pr = probability hit – probability false alarm). While recognition performance did not differ between stimuli classes [mean Prbuilding = 0.40 (SD = 0.14) versus Prlandscape = 0.43 (SD = 0.15), t15 = 1.019, n.s.], it was well above chance level [mean Pr = 0.41 (SD = 0.13), t15 = 13.01; P < 0.0001]. Collapsing across both stimuli classes (building and landscape), we obtained a sufficient number of trials for each response category to reach an adequate contrast-to-noise ratio for our ER-fMRI analyses (78–153 trials per subject for hits, 58–113 for misses, 93–170 for correct rejections and 52–97 for false alarms).

An ANOVA comparing reaction times (Table 1) for hits, misses, correct rejections, and false alarms revealed a reliable effect of response category [F(3,45) = 8.42, P < 0.005]. Post-hoc paired-sample t-tests showed that reactions to correctly identified old items were faster than incorrect reactions to old items (t15 = 3.78, P < 0.005), correct reactions to new items (t15 = 3.21, P < 0.01) and incorrect reactions to new items (t15 = 5.99, P < 0.0001). All other post-hoc tests did not reveal any reliable difference (max t15 = 1.51, n.s.).

Imaging Data

In an exploratory analysis, we directly compared encoding activity to photographs showing either buildings or landscapes. Processing of building stimuli compared to processing of landscape stimuli showed small bilateral, left lateralized activations in superior temporal areas [Talairach and Tournoux (1988) coordinates: x = 52, y = 8, z = –16 and x = –44, y = –4, z = –16], while processing of landscapes as compared to buildings showed more activity in a small area of the right middle frontal gyrus [Talairach and Tournoux (1988) coordinates: x = 40, y = 20, z = 40). These findings may indicate a slightly higher degree of verbal coding for buildings and non-verbal visual-perceptual coding for landscapes (Kelley et al., 1998). Given, however that these small differential effects are not of primary interest for the purpose of our study and that there is no difference in recognition performance, both stimuli classes were pooled together to increase statistical power for all further analyses.

Subsequent Memory Effect

Initially, we sought to verify prior results regarding brain regions involved in successful formation of new declarative memories. Addressing this question requires a comparison between learning events that lead to the successful and unsuccessful formation of memories. As in previous studies, we acquired brain responses to each item during study and conducted contrasts to compare events that were remembered and those that were forgotten as measured by the subsequent recognition memory test during the second experimental run. Figure 1 and Table 2 show brain regions that exhibit significantly more activity to subsequently recognized than forgotten items. In line with previous findings, these encoding areas comprise bilateral fusiform and parahippocampal areas as well as areas in the left basal and lateral frontal cortex [Brodmann Area (BA) 45, 47]. Additionally, we found an activation in the left parietal lobe located in the angular gyrus (BA 39). All activations appear to be more pronounced in the left than the right hemisphere.

Negative Subsequent Memory Effect

In addition to areas predicting subsequent memory by an increase of activation, we intended to identify areas in which a decrease of activation is associated with subsequent memory. Table 2 shows areas that exhibit significantly more activity during learning for subsequently forgotten as opposed to subsequently remembered items. For this contrast we found activations in right medial parietal cortex (BA 7) and in left posterior cingulate cortex.

Repetition Priming Effect

Repetition priming refers to an implicit memory phenomenon in which repeatedly presented items are processed more efficiently, most often accompanied by weaker brain responses to old as opposed to new items. Figure 2A and Table 2 show brain regions exhibiting a decrease of activation to previously seen stimuli as opposed to new ones. As expected, we found activations in bilateral middle occipital gyri (BA 18/19). Moreover, there is an activation in left lingual gyrus.

Old/New Effect

Figure 2B and Table 2 show five brain regions that exhibit more neural activity for correctly identified old items (hits) than for correctly identified new items (correct rejections): (i) a medial-superior frontal area including the superior frontal gyrus medially and laterally in the dorsal-lateral prefrontal cortex (DLPFC), the anterior cingulate and pre- as well as supplementary motor areas (BA 6, 32); (ii) an area in the right superior frontal gyrus (anterior prefrontal cortex: APFC, BA 10); (iii) bilateral, left lateralized areas in the parietal lobe within BA 7 and 40; (iv) an area in the left insula cortex (BA 13); (v) and, finally, bihemispheric activations in the cerebellum.

Positive Recognition Effect

To explore in more detail brain areas engaged in successful memory retrieval, we examined the difference in brain responses to correctly recognized old items (hits) and old items misclassified as new (misses). This contrast (Fig. 3A and Table 2) shows some overlap with activations seen in the old/new contrast regarding gross anatomy, but with the following differences: (i) the right prefrontal activation is centered in the middle frontal gyrus (BA 6) instead of area BA 10; (ii) contrary to the cerebellar old/new effect, the cerebellar recognition effect appears to be stronger and more pronounced at midline structures like the vermis, the intermediate cerebellar hemispheres and the tonsils and it is extended to the relay station for cerebellar afferents, the pons; and (iii) there is an activation of the right angular gyrus, which is not seen in the old/new effect. Nevertheless, the overall location of activations in prefrontal, parietal and cerebellar areas is not entirely different from the old/new effect.

Negative Recognition Effect

Negative recognition effects were obtained by comparing brain responses to misses with responses to hits. As described in the introduction, repetition priming contaminates this effect to a lesser degree. Applying the same minimal cluster size as used for all other contrasts did not lead to a reliable negative recognition effect. However, considering clusters consisting of five voxels or more reveals a left anterior MTL activation (Fig. 3B, Table 2), which is exactly in line with our hypothesis based on electrophysiological findings in humans (Smith et al., 1986) and animals (Brown and Aggleton, 2001). The activation is centered in the anterior parahippocampal gyrus, but, as can be seen in Figure 3B, the activation might extend into the hippocampus.

The Subsequent Memory Effect Inclusively Masked by the Positive Recognition Effect

To identify brain areas showing increased activity for both successful declarative memory formation and retrieval, we masked the subsequent memory effect by the positive recognition effect. Figure 4A and Table 2 show regions that are activated by these two contrasts. In line with our hypotheses regarding the critical role of the inferior and medial temporal lobe we identified bilateral activations in the anterior half of the inferior temporal cortex. In both hemispheres, this area reaches the depth of the collateral sulcus, which is covered by perirhinal cortex (Amaral and Insausti, 1990). Additionally, we identified a major activation in the cerebellar vermis and its afferent relay station, the pons. Further activations are located in left fusiform gyrus, right angular gyrus and in bilateral cerebellar hemispheres.

The Subsequent Memory Effect Inclusively Masked by the Negative Recognition Effect

Inclusively masking the subsequent memory effect by the negative recognition effect allows the identification of brain areas associated with activity increases during successful memory formation and activity decreases during successful memory retrieval. Figure 4B and Table 2 show brain areas exhibiting such a pattern of reactivity. Again, in line with our hypotheses we revealed an anterior MTL activation including the left hippocampus, but centered in the left parahippocampal gyrus.

Discussion

Memory Formation

Replicating almost all earlier findings, we revealed subsequent memory effects in a fusiform/parahippocampal and two left inferior frontal areas, one in the posterior and one in the anterior aspect of the inferior frontal gyrus (Brewer et al., 1998; Wagner et al., 1998; Kirchhoff et al., 2000; Davachi et al., 2001; Otten et al., 2001; Otten and Rugg, 2001a; Strange et al., 2002). The additional subsequent memory effects in the parietal lobe and the cerebellar hemisphere were previously less often reported (Davachi et al., 2001; Otten and Rugg, 2001a). However, the left lateralization of the subsequent memory effects found here is not exactly in line with other studies also using picture stimuli (Brewer et al., 1998; Kirchhoff et al., 2000). It might be explained by the additional use of verbal codes for picture details (Kelley et al., 1998; Opitz et al., 2000). Regardless, our results confirm that prefrontal and medial temporal areas are involved in declarative memory formation, where prefrontal cortex may execute working memory operations associated with maintenance, selection and organization of incoming information (Wagner, 1999; Fletcher and Henson, 2001) and the MTL may execute a rather specific operation of declarative memory formation in the hippocampus and a subordinate support operation in the parahippocampal region. This support operation may make semantic representations of each study item available in the service of comprehension, semantic-associative processing, and memory formation (Nobre and McCarthy, 1995; Fernández et al., 2002).

The negative subsequent memory contrast, more activity for subsequently forgotten than subsequently remembered items, revealed only two small clusters of voxels in the precuneus and the cingulate gyrus. The location of these activations is roughly congruent with activations described in the initial reports of this effect (Otten and Rugg, 2001b; Wagner and Davachi, 2001). These positive correlates of forgetting have been interpreted as related to task-appropriate and task-inappropriate allocation of neurocognitive resources away from the process leading to effective memory formation (Otten and Rugg, 2001b; Wagner and Davachi, 2001).

Recognition

The old/new contrast revealed major activations in the parietal lobe, in frontal midline structures (anterior cingulate and the superior frontal gyrus), the left insula, the right anterior aspect of the superior frontal gyrus and in both cerebellar hemispheres with a left hemispheric dominance. As intended, these findings replicate earlier ER-fMRI findings that suggest distributed cerebral and cerebellar brain regions participating in recognition memory (Henson et al., 1999; Konishi et al., 2000; McDermott et al., 2000; Cabeza et al., 2001; Donaldson et al., 2001a,b). Among these regions, the midline structures activated (anterior cingulate and the superior frontal gyrus) might control subject responses by evaluating stimulus representations restored in the parietal lobe (Buckner et al., 1996; Fletcher et al., 1996). Especially with the large number of stimuli used here, a high degree of interference or response competition makes an effective control of response selection and inhibition necessary (Carter et al., 1998; Braver et al., 2001; Potts and Tucker, 2001; Stern et al., 2001; Levy and Anderson, 2002). Together with the left prefrontal subsequent memory effect described above, the right anterior prefrontal activation is fully in accord with the hemispheric encoding/retrieval asymmetry (HERA) model of a prefrontal encoding and retrieval asymmetry as proposed initially by Tulving et al. (1994). The right anterior prefrontal activation might correlate with postretrieval monitoring processes and not with the actual process of memory retrieval (Rugg et al., 1996; Schacter et al., 1997; Buckner et al., 1998). The old/new effects in cerebellar hemispheres indicate that the cerebellum plays a role in memory retrieval (Bäckman et al., 1997; Cabeza et al., 1997; Andreasen et al., 1999). The implication of this finding will be discussed below, interpreted in the context of the areas involved in both memory formation and retrieval.

The newly introduced positive recognition contrast (i.e. more activity for hits than misses) revealed activations in the frontal and parietal lobe that are close to activations revealed by the old/new contrast, but without direct overlap. The cerebellar activation is compared to the old/new effect more centered at midline structures (i.e. vermis, intermediate cerebellar hemispheres and tonsils) and extended to pontine areas where input from prefrontal, parietal and temporal cortices is relayed to the cerebellum (Schmahmann, 1996). The failure to find exact overlap between the positive recognition- and the old/new effect must, like all null results, be treated with caution. This is especially so, given that the power to detect a recognition effect was lower than the power to detect an old/new effect, a consequence of fewer old misses than new correct rejections. Nevertheless, our findings seem to support the view that frontal, parietal and cerebellar regions are involved in the successful recovery of declarative memories during a recognition memory task.

The small negative recognition effect may indicate that less activity in the anterior MTL is related to recognition success. This finding is in line with our hypothesis and electrophysiological studies (Smith et al., 1986; Riches et al., 1991; Miller and Desimone, 1994; Brown and Aggleton, 2001). It is unlikely that this effect is solely based on repetition priming, because both classes of items have been encountered once before. However, since our study design does not provide a behavioral measure of repetition priming, we are unable to test stochastic independence between primed and recognized items. Though, the location of priming effects in occipital areas only (Fig. 2B) makes a repetition priming account for the negative recognition effect in the anterior MTL highly unlikely. Nevertheless, there seem to be alternative interpretations for the negative recognition effect: old items misclassified as new could be re-encoded during the test phase leading to an encoding related activity increase (Buckner et al., 2001), or the subjects’ new-decision could be accompanied by an activity increase related to novelty detection (Tulving and Kroll, 1995; Tulving et al., 1996). The first alternative interpretation is not mutually exclusive with the recognition account (see below) and the latter interpretation seems to be less plausible, because a reversed old/new contrast with more statistical power (more items) did not show any activation in the anterior MTL (data not shown).

Given our study design, we cannot dissociate between subprocesses within recognition – whether an activation is related to recollection or familiarity (Mandler, 1980). This issue could be further evaluated by a study design including for instance a source memory judgment (Cansino et al., 2002). However, following Brown and Aggleton (2001) or Brown and Bashir (2002), the negative recognition effect in the anterior MTL might rather support a familiarity-based decision than an actual recollective experience (Henson et al., 2003). It may reflect a process enabling recognition by more efficient processing of recognized stimuli with reduced neural activity during an active memory search (Jiang et al., 2000), or by a neural activity increase in the presence of novel stimuli or old stimuli incorrectly classified as new (Brown and Bashir, 2002).

Memory Formation and Recognition

The largest activation clusters of the subsequent memory effect masked by the positive recognition effect are located in the inferior and anterior medial temporal lobe as well as cerebellar and pontine regions. When masking the subsequent memory effect by the negative recognition effect, activations are located in the MTL

Thus, the anterior inferior temporal cortex including the anterior parahippocampal region seems to be conjointly involved in both successful encoding and recognition. This finding confirms suggestions based on across-study comparisons of electrophysiological findings in epilepsy patients (Smith et al., 1986; Fernández et al., 1999, 2001, 2002). Such a module has originally been described on the basis of electrical recordings in non-human primates showing that this brain area is sensitive to both object encoding and object recognition (Desimone et al., 1984; Riches et al., 1991; Miller and Desimone, 1994). During recognition, the neural representation of each test stimulus, i.e. a unique pattern of activation that is evoked by a visually perceived item during recognition, may be matched with stored representations previously formed locally during encoding. Moreover, the inferior and medial temporal cortex is ideally located for this efficient pattern matching, because it is the final route of the ventral visual pathway, providing integrated visual and semantic information (Ungerleider and Mishkin, 1985; Haxby et al., 1991; Nobre and McCarthy, 1995; Büchel et al., 1998; Lerner et al., 2001).

Our findings suggest an important role of the cerebellum and its afferent relay station, the pons, in declarative memory. Functional imaging studies provide mounting evidence that the cerebellum coordinates diverse aspects of cognitive processes (for a review, see Desmond and Fiez, 1998). Up to now, however, it is unclear whether the cerebellum provides domain-general computations supporting diverse cognitive operations or different operations with specific roles in particular cognitive domains. Several proposals have been made for a general operation, including a central timing processor (Keele and Ivry, 1990) for sequential parsing of temporally complex material (Llinas, 1974; De Zeeuw et al., 1998). Thach (1998) proposed that cerebellar processing entails stimulus–response linkage by grouping single-response elements into larger task adequate combinations. The cerebellum also seems to be involved in processes contributing specifically to learning and memory. It is not only critically involved in basic delay conditioning, where it is the locus of memory formation, consolidation, and storage (Thompson and Kim, 1996; Attwell et al., 2002), it is also involved in spatial learning and memory (Pellegrino and Altman, 1979; Lalonde and Botez, 1990; Goodlett et al., 1992). Humans with acquired cerebellar lesions have, however, only minor deficits in declarative memory (Schmahmann, 1998). In imaging studies of declarative memory, cerebellar activations were rather obtained during retrieval than encoding tasks (Desmond and Fiez, 1998), suggesting that a cortical-cerebellar network self-initiates and monitors conscious retrieval (Bäckman et al., 1997; Andreasen et al., 1999) or that the cerebellum generates candidate responses during a search and selection process (Cabeza et al., 1997; Desmond et al., 1998). Our data show that the cerebellum participates in both memory formation and retrieval. However, the fact that cerebellar lesions cause only minor deficits in declarative memory suggests that the cerebellum is not directly involved in storage and retrieval operations. It might rather support mnemonic operations by providing a temporal structure for a coherent episode.

In conclusion, by replicating ER-fMRI studies investigating either memory formation or recognition we have provided within-study confirmation for brain areas involved in two fundamental mnemonic operations: either the formation or the retrieval of declarative memories. Based on this empirical foundation, we have described for the first time brain regions supporting successful memory retrieval by both activity increases (frontal, parietal, cerebellar areas) and decreases (anterior MTL). Finally, we have initially identified within subjects and within one experiment inferior- and medial-temporal as well as cerebellar areas supporting both memory formation and retrieval. Such integrated modules may re-use stored representations formed locally during encoding for efficient matching operations during recognition.

Acknowledgements

We thank Doug Davidson, Karl Magnus Petersson, Indira Tendolkar and Miranda van Turennout for instructive comments on the manuscript and Karsten Specht for technical advice in data acquisition and analysis. G.F. is supported by BONFOR, the intramural research support program and the German Research Council (DFG Fe479/4-1; SFB-TR 3/A4).

Figure 1. Subsequent memory effect. Regions activated more in case of successful as opposed to unsuccessful memory formation during encoding. The activation map (P < 0.005, uncorrected; minimal cluster size 10 voxels) is shown overlaid onto a canonical brain rendered in three dimensions. Specific activations are additionally shown superimposed onto selected coronal slices of the mean high-resolution T1-weighted volume. Slices are numbered according to coordinates of Talairach and Tournoux (1988). IFGa, anterior aspect of the inferior frontal gyrus; IFGp, posterior aspect of the inferior frontal gyrus; PHG, parahippocampal gyrus.

Figure 1. Subsequent memory effect. Regions activated more in case of successful as opposed to unsuccessful memory formation during encoding. The activation map (P < 0.005, uncorrected; minimal cluster size 10 voxels) is shown overlaid onto a canonical brain rendered in three dimensions. Specific activations are additionally shown superimposed onto selected coronal slices of the mean high-resolution T1-weighted volume. Slices are numbered according to coordinates of Talairach and Tournoux (1988). IFGa, anterior aspect of the inferior frontal gyrus; IFGp, posterior aspect of the inferior frontal gyrus; PHG, parahippocampal gyrus.

Figure 2. Repetition priming effect (A) and old/new effect (B). Regions activated more for new as opposed to old stimuli during recognition (A). Regions activated more for hits as opposed to correct rejections during recognition (B). Activation maps (P < 0.005, uncorrected; minimal cluster size 10 voxels) are shown overlaid onto a canonical brain rendered in three dimensions. Specific activations are additionally shown superimposed onto selected coronal slices of the mean high-resolution T1-weighted volume. Slices are numbered according to coordinates of Talairach and Tournoux (1988). ACG, anterior aspect of the cingulate gyrus; CH, cerebellar hemisphere; LG, lingual gyrus; MOG, middle occipital gyrus; PCG, precentral gyrus; SFG, superior frontal gyrus; SPL, superior parietal lobule.

Figure 2. Repetition priming effect (A) and old/new effect (B). Regions activated more for new as opposed to old stimuli during recognition (A). Regions activated more for hits as opposed to correct rejections during recognition (B). Activation maps (P < 0.005, uncorrected; minimal cluster size 10 voxels) are shown overlaid onto a canonical brain rendered in three dimensions. Specific activations are additionally shown superimposed onto selected coronal slices of the mean high-resolution T1-weighted volume. Slices are numbered according to coordinates of Talairach and Tournoux (1988). ACG, anterior aspect of the cingulate gyrus; CH, cerebellar hemisphere; LG, lingual gyrus; MOG, middle occipital gyrus; PCG, precentral gyrus; SFG, superior frontal gyrus; SPL, superior parietal lobule.

Figure 3. Recognition effects. Regions activated more for hits as opposed to misses (positive recognition effect, A). Regions activated less for hits as opposed to misses (negative recognition effect, B). For (A), the activation maps (P < 0.005 uncorrected; minimal cluster size 10 voxels) is shown overlaid onto a canonical brain rendered in three dimensions. Specific activations are additionally shown superimposed onto selected coronal slices of the mean high-resolution T1-weighted volume. For (B), all activations are shown on two slices (P < 0.005 uncorrected; minimal cluster size five voxels). Slices are numbered according to coordinates of Talairach and Tournoux (1988). AG, angular gyrus; MFG, middle frontal gyrus; Po, pons; RC/Hi, rhinal cortex/hippocampus; VE, vermis.

Figure 3. Recognition effects. Regions activated more for hits as opposed to misses (positive recognition effect, A). Regions activated less for hits as opposed to misses (negative recognition effect, B). For (A), the activation maps (P < 0.005 uncorrected; minimal cluster size 10 voxels) is shown overlaid onto a canonical brain rendered in three dimensions. Specific activations are additionally shown superimposed onto selected coronal slices of the mean high-resolution T1-weighted volume. For (B), all activations are shown on two slices (P < 0.005 uncorrected; minimal cluster size five voxels). Slices are numbered according to coordinates of Talairach and Tournoux (1988). AG, angular gyrus; MFG, middle frontal gyrus; Po, pons; RC/Hi, rhinal cortex/hippocampus; VE, vermis.

Figure 4. Subsequent memory masked inclusively by recognition effects. Regions activated more during successful as opposed to unsuccessful memory formation and more for hits as opposed to misses (subsequent memory effect and positive recognition effect, A). Regions activated more during successful as opposed to unsuccessful memory formation and less for hits as opposed to misses (subsequent memory effect and negative recognition effect, B). For (A), the activation maps (P < 0.017 uncorrected; minimal cluster size 10 voxels) is shown overlaid onto a canonical brain rendered in three dimensions. Specific activations are additionally shown superimposed onto selected coronal slices of the mean high-resolution T1-weighted volume. For (B), all activations are shown on two slices (P < 0.017 uncorrected; minimal cluster size five voxels). Slices are numbered according to coordinates of Talairach and Tournoux (1988). AG, angular gyrus; FG, fusiform gyrus; ITG, inferior temporal gyrus; RC, rhinal cortex; Hi, hippocampus; Po, pons; Ve, vermis.

Figure 4. Subsequent memory masked inclusively by recognition effects. Regions activated more during successful as opposed to unsuccessful memory formation and more for hits as opposed to misses (subsequent memory effect and positive recognition effect, A). Regions activated more during successful as opposed to unsuccessful memory formation and less for hits as opposed to misses (subsequent memory effect and negative recognition effect, B). For (A), the activation maps (P < 0.017 uncorrected; minimal cluster size 10 voxels) is shown overlaid onto a canonical brain rendered in three dimensions. Specific activations are additionally shown superimposed onto selected coronal slices of the mean high-resolution T1-weighted volume. For (B), all activations are shown on two slices (P < 0.017 uncorrected; minimal cluster size five voxels). Slices are numbered according to coordinates of Talairach and Tournoux (1988). AG, angular gyrus; FG, fusiform gyrus; ITG, inferior temporal gyrus; RC, rhinal cortex; Hi, hippocampus; Po, pons; Ve, vermis.

Table 1


 Mean recognition performance and reaction times (RT) with their standard deviations (SD)

 Old  New  
 Hits Misses  Correct rejections False alarms Uncertain 
       
Number  130   66   140   50   95 
SD   28   23    23   19   62 
       
RT (ms) 1382 1479  1462 1445 1727 
SD  194  214   187  196  103 
 Old  New  
 Hits Misses  Correct rejections False alarms Uncertain 
       
Number  130   66   140   50   95 
SD   28   23    23   19   62 
       
RT (ms) 1382 1479  1462 1445 1727 
SD  194  214   187  196  103 
Table 2


 Activation peaks with their localization, significance level and the size of the respective activation cluster (number of voxels)

Effect Anatomical region  BA Coordinates   t-value No. of voxels 
    x y z   
Subsequent memory effect Left angular g.  39 –28 –72  40 5.27  23 
 Left fusiform g.  20 –44 –48 –28 6.15  85 
 Right fusiform g.  37  24 –56 –16 4.31  37 
 Left. parahippocampal g. PHG  35 –20  –8 –36 4.15  11 
 Right parahippocampal g. PHG 35  36 –12  36 6.60  27 
 Left inferior frontal g. IFGa 47 –44  44 –12 4.56  36 
 Left inferior frontal g. IFGp 45 –40  12  28 6.60  40 
         
Negative subsequent memory effect Right precuneus   7   8 –48  52 3.94  18 
 Left cingulate g.  24 –16   4  36 5.30  11 
         
Repetition priming effect Left middle occipital g.  MOG 19 –36 –92   8 5.15  17 
 Left middle occipital g.  MOG 18 –20 –92  20 4.47  14 
 Right middle occipital g.  MOG 19  28 –84   0 3.68  12 
 Left lingual g. LG 19 –32 –64  –4 4.81  15 
         
Old/new effect Left superior parietal l. SPL  7 –32 –64  52 5.56 176 
 Left inferior parietal l.   40 –56 –32  40 5.05  19 
 Right inferior parietal l.   40  36 –40  40 5.26  53 
 Right fusiform g.   37  40 –48 –16 5.37  10 
 Left superior temporal g.   22 –60 –48  16 3.99  10 
 Left cerebellum CH  –20 –56 –36 6.32  36 
 Left cerebellum CH  –44 –60 –48 4.68  39 
 Right cerebellum CH   20 –60 –36 4.18  22 
 Right cerebellum CH   48 –64 –40 4.08  12 
 Left insula   13 –36  –4  16 4.09  13 
 Left precentral g.  PCG  6 –40   0  36 5.18  22 
 Left superior frontal g. SFG   6   0  16  56 6.03  94 
 Right cingulate g. ACG 32  –4  36  28 7.08  34 
 Right superior frontal g.  SFG 10  32  60  16 5.53  17 
 Right inferior frontal g.   47  32  20 –16 4.15  13 
 Right middle frontal g.  32  12  12  44 3.86  11 
 Left cingulate g.   23  –4 –12  24 4.79  27 
         
Positive recognition effect Right angular g.  AG 39  40 –60  32 3.73  11 
 Vermis, cerebellum Ve    0 –60 –20 4.82  42 
 Left pons Po  –16 –48 –24 5.63  52 
 Right middle frontal g. MFG  6  16 –12   56 4.86  13 
 Left cerebellum   –20 –72 –28 4.18  15 
         
Negative recognition effect Left parahippocampal g. RC/Hi 35 –24 –20 –16 3.77   6 
         
The subsequent memory masked inclusively by the positive recognition effect Left cerebellum   –12 –64 –28 3.76  23 
 Vermis, cerebellum  Ve    0 –60 –20 3.21  23 
 Right angular g. AG 39  40 –60  32 3.85  14 
 Left pons Po  –28 –32 –28 3.83  34 
 Left fusiform g. FG 20 –36 –60 –24 4.23  30 
 Left inferior temporal g. ITG 20 –56 –16 –24 3.81  29 
 Left parahippocampal g. RC 20 –36  –8 –24 2.51  12 
 Right parahippocampal g. RC 20  40   0 –36 3.56  12 
         
The subsequent memory masked inclusively by the negative recognition effect Left parahippocampal g.  RC/Hi 28 –24 –16 –16 2.99  17 
 Left cerebellum   –44 –48 –28 5.37  12 
Effect Anatomical region  BA Coordinates   t-value No. of voxels 
    x y z   
Subsequent memory effect Left angular g.  39 –28 –72  40 5.27  23 
 Left fusiform g.  20 –44 –48 –28 6.15  85 
 Right fusiform g.  37  24 –56 –16 4.31  37 
 Left. parahippocampal g. PHG  35 –20  –8 –36 4.15  11 
 Right parahippocampal g. PHG 35  36 –12  36 6.60  27 
 Left inferior frontal g. IFGa 47 –44  44 –12 4.56  36 
 Left inferior frontal g. IFGp 45 –40  12  28 6.60  40 
         
Negative subsequent memory effect Right precuneus   7   8 –48  52 3.94  18 
 Left cingulate g.  24 –16   4  36 5.30  11 
         
Repetition priming effect Left middle occipital g.  MOG 19 –36 –92   8 5.15  17 
 Left middle occipital g.  MOG 18 –20 –92  20 4.47  14 
 Right middle occipital g.  MOG 19  28 –84   0 3.68  12 
 Left lingual g. LG 19 –32 –64  –4 4.81  15 
         
Old/new effect Left superior parietal l. SPL  7 –32 –64  52 5.56 176 
 Left inferior parietal l.   40 –56 –32  40 5.05  19 
 Right inferior parietal l.   40  36 –40  40 5.26  53 
 Right fusiform g.   37  40 –48 –16 5.37  10 
 Left superior temporal g.   22 –60 –48  16 3.99  10 
 Left cerebellum CH  –20 –56 –36 6.32  36 
 Left cerebellum CH  –44 –60 –48 4.68  39 
 Right cerebellum CH   20 –60 –36 4.18  22 
 Right cerebellum CH   48 –64 –40 4.08  12 
 Left insula   13 –36  –4  16 4.09  13 
 Left precentral g.  PCG  6 –40   0  36 5.18  22 
 Left superior frontal g. SFG   6   0  16  56 6.03  94 
 Right cingulate g. ACG 32  –4  36  28 7.08  34 
 Right superior frontal g.  SFG 10  32  60  16 5.53  17 
 Right inferior frontal g.   47  32  20 –16 4.15  13 
 Right middle frontal g.  32  12  12  44 3.86  11 
 Left cingulate g.   23  –4 –12  24 4.79  27 
         
Positive recognition effect Right angular g.  AG 39  40 –60  32 3.73  11 
 Vermis, cerebellum Ve    0 –60 –20 4.82  42 
 Left pons Po  –16 –48 –24 5.63  52 
 Right middle frontal g. MFG  6  16 –12   56 4.86  13 
 Left cerebellum   –20 –72 –28 4.18  15 
         
Negative recognition effect Left parahippocampal g. RC/Hi 35 –24 –20 –16 3.77   6 
         
The subsequent memory masked inclusively by the positive recognition effect Left cerebellum   –12 –64 –28 3.76  23 
 Vermis, cerebellum  Ve    0 –60 –20 3.21  23 
 Right angular g. AG 39  40 –60  32 3.85  14 
 Left pons Po  –28 –32 –28 3.83  34 
 Left fusiform g. FG 20 –36 –60 –24 4.23  30 
 Left inferior temporal g. ITG 20 –56 –16 –24 3.81  29 
 Left parahippocampal g. RC 20 –36  –8 –24 2.51  12 
 Right parahippocampal g. RC 20  40   0 –36 3.56  12 
         
The subsequent memory masked inclusively by the negative recognition effect Left parahippocampal g.  RC/Hi 28 –24 –16 –16 2.99  17 
 Left cerebellum   –44 –48 –28 5.37  12 

Coordinates are listed in Talairach and Tournoux (1988) atlas space. BA is the Brodmann area nearest to the coordinate and should be considered approximate (g., gyrus).

References

Amaral DG, Insausti R (
1990
) Hippocampal formation. In: The human nervous system (Paxinos G, ed.), pp.
711
–755. San Diego, CA: Academic Press.
Andreasen NC, O‘Leary DS, Paradiso S, Cizadlo T, Arndt S, Watkins GL, Ponto LL, Hichwa RD (
1999
) The cerebellum plays a role in conscious episodic memory retrieval.
Hum Brain Mapp
 
8
:
226
–234.
Attwell PJ, Cooke SF, Yeo CH (
2002
) Cerebellar function in consolidation of a motor memory.
Neuron
 
34
:
1011
–1020.
Bäckman L, Almkvist O, Andersson J, Nordberg A, Winblad B, Reineck R, Långström B (
1997
) Brain activation in young and older adults during implicit and explicit retrieval.
J Cogn Neurosci
 
9
:
378
–391.
Braver TS, Barch DM, Gray JR, Molfese DL, Snyder A (
2001
) Anterior cingulate cortex and response conflict: effects of frequency, inhibition and errors.
Cereb Cortex
 
11
:
825
–836.
Brewer JB, Zhao Z, Desmond JE, Glover GH, Gabrieli JDE (
1998
) Making memories: brain activity that predicts how well visual experience will be remembered.
Science
 
281
:
1185
–1187.
Brown MW, Aggleton JP (
2001
) Recognition memory: what are the roles of the perirhinal cortex and hippocampus.
Nat Rev Neurosci
 
2
:
51
–61.
Brown MW, Bashir ZI (
2002
) Evidence concerning how neurons of the perirhinal cortex may effect familiarity discrimination.
Philos Trans R Soc Lond B Biol Sci
 
357
:
1083
–1095.
Brown MW, Wilson FA, Riches IP (
1987
) Neuronal evidence that inferomedial temporal cortex is more important than hippocampus in certain processes underlying recognition memory.
Brain Res
 
409
:
158
–162.
Büchel C, Price C, Friston K (
1998
) A multimodal language region in the ventral visual pathway.
Nature
 
394
:
274
–276.
Buckner RL, Koutstaal W (
1998
) Functional neuroimaging studies of encoding, priming, and explicit memory retrieval.
Proc Natl Acad Sci USA
 
95
:
891
–898.
Buckner RL, Raichle ME, Miezin FM, Petersen SE (
1996
) Functional anatomic studies of memory retrieval for auditory words and visual pictures.
J Neurosci
 
16
:
6219
–6235.
Buckner RL, Koutstaal W, Schacter DL, Wagner AD, Rosen BR (
1998
) Functional-anatomic study of episodic retrieval using fMRI. I. Retrieval effort versus retrieval success.
Neuroimage
 
7
:
151
–162.
Buckner RL, Wheeler ME, Sheridan MA (
2001
) Encoding processes during retrieval tasks.
J Cogn Neurosci
 
13
:
406
–415.
Cabeza R, Kapur S, Craik FIM, McIntosh AR, Houle S, Tulving E (
1997
) Functional neuroanatomy of recall and recognition: a PET study of episodic memory.
J Cogn Neurosci
 
9
:
254
–265.
Cabeza R, Rao SM, Wagner AD, Mayer AR, Schacter DL (
2001
) Can medial temporal lobe regions distinguish true from false? An event-related functional MRI study of veridical and illusory recognition memory.
Proc Natl Acad Sci USA
 
98
:
4805
–4810.
Cansino S, Maquet P, Dolan RJ, Rugg MD (
2002
) Brain activity underlying encoding and retrieval of source memory.
Cereb Cortex
 
12
:
1048
–1056.
Carter CS, Braver TS, Barch DM, Botvinick MM, Noll D, Cohen JD (
1998
) Anterior cingulate cortex, error detection, and the online monitoring of performance.
Science
 
280
:
747
–749.
Cohen NJ, Squire LR (
1980
) Preserved learning and retention of pattern analyzing skill in amnesia: dissociation of knowing how and knowing that.
Science
 
210
:
207
–210.
Dale AM, Buckner RL (
1997
) Selective averaging of rapidly presented individual trials using fMRI.
Hum Brain Mapp
 
5
:
329
–340.
Davachi L, Maril A, Wagner AD (
2001
) When keeping in mind supports later bringing to mind: neural markers of phonological rehearsal predict subsequent remembering.
J Cogn Neurosci
 
13
:
1059
–1070.
Desimone R, Albright TD, Gross CG, Bruce C (
1984
) Stimulus-selective properties of inferior temporal neurons in the macaque.
J Neurosci
 
4
:
2051
–2062.
Desmond JE, Fiez JA (
1998
) Neuroimaging studies of the cerebellum: language, learning and memory.
Trends Cogn Sci
 
2
:
355
–362.
Desmond JE, Gabrieli JDE, Glover GH (
1998
) Dissociation of frontal and cerebellar activity in a cognitive task: evidence for a distinction between selection and search.
Neuroimage
 
7
:
368
–376.
De Zeeuw CI, Simpson JI, Hoogenraad CC, Galjart N, Koekkoek SK, Ruigrok TJ (
1998
) Microcircuitry and function of the inferior olive.
Trends Neurosci
 
21
:
391
–400.
Donaldson DI, Buckner RL (
1999
) Trying versus succeeding: event-related designs dissociate memory processes.
Neuron
 
22
:
412
–414.
Donaldson DI, Petersen SE, Buckner RL (
2001
) Dissociating memory retrieval processes using fMRI: evidence that priming does not support recognition memory.
Neuron
 
31
:
1047
–1059.
Donaldson DI, Petersen SE, Ollinger JM, Buckner RL (
2001
) Dissociating state and item components of recognition memory using fMRI.
Neuroimage
 
13
:
129
–142.
Fernández G, Effern A, Grunwald T, Pezer N, Lehnertz K, Dümpelmann M, Van Roost D, Elger CE (
1999
) Real-time tracking of memory formation in the human rhinal cortex and hippocampus.
Science
 
285
:
1582
–1585.
Fernández G, Heitkemper P, Grunwald T, Van Roost D, Urbach H, Pezer N, Lehnertz K, Elger CE (
2001
) An inferior temporal stream for word processing with integrated mnemonic function.
Hum Brain Mapp
 
14
:
251
–260.
Fernández G, Klaver P, Fell J, Grunwald T, Elger CE (
2002
) Human declarative memory formation: segregating rhinal and hippocampal contributions.
Hippocampus
 
12
:
514
–519.
Fletcher PC, Henson RN (
2001
) Frontal lobes and human memory: insights from functional neuroimaging.
Brain
 
124
:
849
–881.
Fletcher PC, Shallice T, Frith CD, Frackowiak RS, Dolan RJ (
1996
) Brain activity during memory retrieval. The influence of imagery and semantic cueing.
Brain
 
119
:
1587
–1596.
Friston KJ, Fletcher P, Josephs O, Holmes A, Rugg MD, Turner R (
1998
) Event-related fMRI: characterizing differential responses.
Neuroimage
 
7
:
30
–40.
Gabrieli JDE (
1998
) Cognitive neuroscience of human memory.
Annu Rev Psychol
 
49
:
87
–115.
Gabrieli JDE, Brewer JB, Desmond JE, Glover GH (
1997
) Separate neural bases of two fundamental memory processes in the human medial temporal lobe.
Science
 
276
:
264
–266.
Goodlett CR, Hamre KM, West JR (
1992
) Dissociation of spatial navigation and visual guidance performance in Purkinje cell degeneration (pcd) mutant mice.
Behav Brain Res
 
47
:
129
–141.
Haxby JV, Grady CL, Horwitz B, Ungerleider LG, Mishkin M, Carson RE, Herscoviych P, Shapiro MB, Rapoport, SI (
1991
) Dissociation of object and spatial visual processing pathways in human extrastriate cortex.
Proc Natl Acad Sci USA
 
88
:
1621
–1625.
Henson RN, Rugg MD, Shallice T, Josephs O, Dolan RJ (
1999
) Recollection and familiarity in recognition memory: an event-related functional magnetic resonance imaging study.
J Neurosci
 
19
:
3962
–3972.
Henson R, Shallice T, Dolan R (
2000
) Neuroimaging evidence for dissociable forms of repetition priming.
Science
 
287
:
1269
–1272.
Henson RNA, Cansino S, Herron JE, Robb WGK, Rugg MD (
2003
) A familiarity signal in human anterior medial temporal cortex?
Hippocampus
 
13
:
259
–262.
Jiang Y, Haxby JV, Martin A, Ungerleider LG, Parasuraman R (
2000
) Complementary neural mechanisms for tracking items in human working memory.
Science
 
287
:
643
–646.
Josephs O, Turner R, Friston K (
1997
) Event-related fMRI.
Hum Brain Mapp
 
5
:
243
–248.
Keele SW, Ivry R (
1990
) Does the cerebellum provide a common computation for diverse tasks? A timing hypothesis.
Ann N Y Acad Sci
 
608
:
179
–207.
Kelley WM, Miezin FM, McDermott KB, Buckner RL, Raichle ME, Cohen NJ, Ollinger JM, Akbudak E, Conturo TE, Snyder AZ, Petersen SE (
1998
) Hemispheric specialization in human dorsal frontal cortex and medial temporal lobe for verbal and nonverbal memory encoding.
Neuron
 
20
:
927
–936.
Kirchhoff BA, Wagner AD, Maril A, Stern CE (
2000
) Prefrontal-temporal circuitry for episodic encoding and subsequent memory.
J Neurosci
 
20
:
6173
–6180.
Konishi S, Wheeler ME, Donaldson DI, Buckner RL (
2000
) Neural correlates of episodic retrieval success.
Neuroimage
 
12
:
276
–286.
Lalonde R, Botez MI (
1990
) The cerebellum and learning processes in animals.
Brain Res
 
Brain Res Rev
 
15
:
325
–332.
Lerner Y, Hendler T, Ben-Basher D, Harel M, Malach R (
2001
) A hierarchical axis of object processing stages in the human visual cortex.
Cereb Cortex
 
11
:
287
–297.
Levy BJ, Anderson MC (
2002
) Inhibitory processes and the control of memory retrieval.
Trends Cogn Sci
 
6
:
299
–305.
Llinas R (
1974
) Eighteenth Bowditch lecture. Motor aspects of cerebellar control.
Physiologist
 
17
:
19
–46.
McCarthy G, Nobre AC, Bentin S, Spencer DD (
1995
) Language-related field potentials in the anterior-medial temporal lobe: I. Intracranial distribution and neural generators.
J Neurosci
 
15
:
1080
–1089.
McDermott KB, Jones TC, Petersen SE, Lageman SK, Roediger HL III (
2000
) Retrieval success is accompanied by enhanced activation in anterior prefrontal cortex during recognition memory: an event-related fMRI study.
J Cogn Neurosci
 
12
:
965
–976.
Mandler D (
1980
) Recognizing: the judgment of previous occurrence.
Psychol Rev
 
87
:
252
–271.
Miller EK, Desimone R (
1994
) Parallel neuronal mechanisms for short-term memory.
Science
 
263
:
520
–522.
Morcom AM, Good CD, Frackowiak RS, Rugg MD (
2003
) Age effects on the neural correlates of successful memory encoding.
Brain
 
126
:
213
–229.
Nobre AC, McCarthy G (
1995
) Language-related field potentials in the anterior-medial temporal lobe: II., Effects of word type and semantic priming.
J Neurosci
 
15
:
090
–1098.
Oldfield RC (
1971
) The assessment and analysis of handedness: the Edinburgh inventory.
Neuropsychologia
 
9
:
97
–113.
Opitz B, Mecklinger A, Friederici AD (
2000
) Functional asymmetry of human prefrontal cortex: encoding and retrieval of verbally and nonverbally coded information.
Learn Mem
 
7
:
85
–96.
Otten LJ, Rugg MD (
2001
) Task-dependency of the neural correlates of episodic encoding as measured by fMRI.
Cereb Cortex
 
11
:
1150
–1160.
Otten LJ, Rugg MD (
2001
) When more means less: neural activity related to unsuccessful memory encoding.
Curr Biol
 
11
:
1528
–1530.
Otten LJ, Henson RN, Rugg MD (
2001
) Depth of processing effects on neural correlates of memory encoding: relationship between findings from across- and within-task comparisons.
Brain
 
124
:
399
–412.
Paller KA, Wagner AD (
2002
) Observing the transformation of experience into memory.
Trends Cogn Sci
 
6
:
93
–102.
Pellegrino LJ, Altman J (
1979
) Effects of differential interference with postnatal cerebellar neurogenesis on motor performance, activity level, and maze learning of rats: a developmental study.
J Comp Physiol Psychol
 
93
:
1
–33.
Potts GF, Tucker DM (
2001
) Frontal evaluation and posterior representation in target detection.
Brain Res Cogn Brain Res
 
11
:
147
–156.
Riches IP, Wilson FA, Brown MW (
1991
) The effects of visual stimulation and memory on neurons of the hippocampal formation and the neighboring parahippocampal gyrus and inferior temporal cortex of the primate.
J Neurosci
 
11
:
1763
–1779.
Rombouts SARB, Barkhof F, Witter MP, Machielsen WCM, Scheltens Ph (
2001
) Anterior medial temporal lobe activation during attempted retrieval of encoded visuospatial scenes: an event-related fMRI study.
Neuroimage
 
14
:
67
–76.
Rugg MD, Henson RNA (
2002
) Episodic memory retrieval: an (event-related) functional neuroimaging perspective. In: The cognitive neuroscience of memory encoding and retrieval (Parker AE, Wilding EL, Bussey T, eds), pp.
7
–28 Hove: Psychology Press.
Rugg MD, Fletcher PC, Frith CD, Frackowiak RS, Dolan RJ (
1996
) Differential activation of the prefrontal cortex in successful and unsuccessful memory retrieval.
Brain
 
119
:
2073
–2083.
Schacter DL, Buckner RL, Koutstaal W, Dale AM, Rosen BR (
1997
) Late onset of anterior prefrontal activity during true and false recognition: an event-related fMRI study.
Neuroimage
 
6
:
259
–269.
Schmahmann JD (
1996
) From movement to thought: anatomic substrates of the cerebellar contribution to cognitive processing.
Hum Brain Mapp
 
4
:
174
–198.
Schmahmann JD (
1998
) Dysmetria of thought: clinical consequences of cerebellar dysfunction on cognition and affect.
Trends Cogn Sci
 
2
:
362
–371.
Shimamura AP (
1985
) Problems with the finding of stochastic independence as evidence for the independence of cognitive processes.
Bull Psychonom Soc
 
23
:
506
–508.
Smith ME, Stapleton JM, Halgren E (
1986
) Human medial temporal lobe potentials evoked in memory and language tasks.
Electroencephalogr Clin Neurophysiol
 
63
:
145
–159.
Stern CE, Sherman SJ, Kirchhoff BA, Hasselmo ME (
2001
) Medial temporal and prefrontal contributions to working memory tasks with novel and familiar stimuli.
Hippocampus
 
11
:
337
–346.
Strange BA, Otten LJ, Josephs O, Rugg MD, Dolan RJ (
2002
) Dissociable human perirhinal, hippocampal, and parahippocampal roles during verbal encoding.
J Neurosci
 
22
:
523
–528.
Talairach J, Tournoux P (
1988
) Co-planar stereotaxic atlas of the human brain. Stuttgart: Thieme.
Thach WT (
1998
) What is the role of the cerebellum in motor learning and cognition?
Trends Cogn Sci
 
2
:
331
–337.
Thompson RF, Kim JJ (
1996
) Memory systems in the brain and localization of a memory.
Proc Natl Acad Sci USA
 
93
:
13438
–13444.
Tulving E, Kroll N (
1995
) Novelty assessment in the brain and long-term memory encoding.
Psychonom Bull Rev
 
2
:
387
–390.
Tulving E, Schacter DL (
1990
) Priming and human memory systems.
Science
 
247
:
301
–306.
Tulving E, Kapur S, Craik FI, Moscovitch M, Houle S (
1994
) Hemispheric encoding/retrieval asymmetry in episodic memory: positron emission tomography findings.
Proc Natl Acad Sci USA
 
91
:
2016
–2020.
Tulving E, Markowitsch HJ, Craik FIM, Habib R, Houle S (
1996
) Novelty and familiarity activations in PET studies of memory encoding and retrieval.
Cereb Cortex
 
6
:
71
–79.
Ungerleider LG, Mishkin M (
1985
) Two cortical visual systems. In: Analysis of visual behavior (Ingle DJ, Goodale MA, Mansfield RJW, eds), pp.
549
–586. Cambridge, MA: MIT Press.
Wagner AD (
1999
) Working memory contributions to human learning and remembering.
Neuron
 
22
:
19
–22.
Wagner AD, Davachi L (
2001
) Cognitive neuroscience: forgetting of things past.
Curr Biol
 
11
:
964
–967.
Wagner AD, Schacter DL, Rotte M, Koutstaal W, Maril A, Dale AM, Rosen BR, Buckner RL (
1998
) Building memories: remembering and forgetting of verbal experiences as predicted by brain activity.
Science
 
281
:
1188
–1191.
Zarahn E, Aguirre G, D’Eposito M (
1997
) A trial-based experimental design for fMRI.
Neuroimage
 
6
:
122
–138.