Abstract
Posterior parietal cortex is frequently activated during episodic memory retrieval but its role during retrieval and its interactions with the hippocampus are not yet clear. In this fMRI study, we investigated the neural bases of recognition memory when study repetitions and retrieval goals were manipulated. During encoding participants studied words either once or three times, and during retrieval they were rewarded more to detect either studied words or new words. We found that (1) dorsal parietal cortex (DPC) was more engaged during detection of items studied once compared to three times, whereas regions in the ventral parietal cortex (VPC) responded more to items studied multiple times; (2) DPC, within a network of brain regions functionally connected to the anterior hippocampus, responded more to items consistent with retrieval goals (associated with high reward); (3) VPC, within a network of brain regions functionally connected to the posterior hippocampus, responded more to items not aligned with retrieval goals (i.e., unexpected). These findings support the hypothesis that DPC and VPC regions contribute differentially to top-down vs. bottom-up attention to memory. Moreover, they reveal a dissociation in the functional profile of the anterior and posterior hippocampi.
Introduction
The ability to recollect specific past events, or episodic memory, depends on the interplay between the bottom–up emergence of stored memory traces and the top–down control of this process according to retrieval goals (e.g., ; ; ; ; ; Scimeca and Badre, 2012; ; ). Functional neuroimaging (fMRI) and neuropsychological research have linked the bottom-up emergence of memories to the hippocampus, which has been characterized as a “stupid” module whose operations, once initiated, run obligatorily (). In contrast, top–down control retrieval processes have been attributed to the prefrontal cortex, deemed necessary to manage encoding and retrieval operations according to retrieval goals, while interacting with the hippocampus and associated medial temporal lobe (MTL) regions (; ; Simons and Spiers, 2003).
More recent research has identified the posterior parietal cortex as another core element of the episodic retrieval network. In fMRI studies, posterior parietal cortex is one of the regions most frequently activated during episodic retrieval, and, critically, it almost always shows greater activity for successfully recognized old items (hits) than successfully rejected new items (correct rejections—CRs), or ‘retrieval success effect’ (Wagner et al., 2005; ; ). Moreover, patients with lesions to the posterior parietal cortex, though not amnesic, do show subtle anterograde and retrograde memory impairments (; ; ,; Simons et al., 2010; ; ; Yazar et al., 2017). The posterior parietal cortex has long been associated with attention – not memory – and, therefore, there have been many attempts to explain the involvement of posterior parietal cortex in episodic memory retrieval (see, for reviews, , ; ; ; ; Sestieri et al., 2017).
The ‘attention to memory’ hypothesis (; ) sprang from the widely agreed observation that the two major divisions of posterior parietal cortex, dorsal parietal cortex (DPC; superior parietal lobule and intraparietal sulcus, roughly corresponding to BA7) and ventral parietal cortex (VPC; angular gyrus and supramarginal gyrus, roughly corresponding to BAs 39 and 40) have different functional profiles. DPC shows greater activity for low than high confidence memory judgments, when the engagement of memory search and top-down monitoring for diagnostic memory content is presumably maximal, whereas VPC is prominently active during the recognition of items of more obvious memory status, such as those accompanied by high confidence or the subjective feeling of recollection (). The ventral/dorsal distinction observed in the memory domain echoes the distinction between the roles of DPC and VPC in attention: DPC supports top–down attention, which enables selection of stimuli based on internal goals, whereas VPC mediates the bottom-up capture of attention following detection of relevant stimuli (; ; ). Consistently, in the “Posner” paradigm, DPC is maximally engaged during the cue period, when participants search for a target, whereas VPC is engaged during target detection, and responds more strongly to invalidly compared to validly cued targets (). According to the attention to memory model, DPC activity maintains retrieval goals, which modulate memory-related activity in the MTLs, whereas VPC activity mediates the change in the locus of attention following detection of relevant memories retrieved by the MTLs. A number of studies have provided empirical support to this model (reviewed in , ). For example, in a cued recognition experiment, DPC was active when participants anticipated a target based on a memory cue, whereas VPC mediated fast detection of memory targets in the absence of cues (; see also ). Moreover, the left angular gyrus of VPC was more active during detection of invalidly vs. validly cued memory contents (; ; ).
Other studies have challenged the attention to memory model on a number of points. pointed out that posterior parietal cortex subregions associated with top–down and bottom–up attention are adjacent but non-overlapping with those associated with episodic retrieval (Sestieri et al., 2017). Additionally, detected multiple response profiles in posterior parietal cortex, of which only some appeared reflective of attention to memory. Therefore, although the dual attention system model was useful to frame the coarse segregation of DPC and VPC memory effects, this framework may not completely capture all the different functional properties of posterior parietal cortex subregions (; Sestieri et al., 2017). Although this debate is beyond the scope of the current article, we acknowledge that the number of studies specifically designed to test the attention to memory model has been limited, and these studies have employed paradigms that resembled attentional paradigms in some respect (e.g., use of cues, violation of expectations; ; ).
To address this issue, the first goal of the current study was to test the attention to memory model using a standard recognition memory paradigm. We did so by manipulating two factors deemed to differentially affect bottom–up and top–down attention to memory: study repetitions and retrieval goals. During encoding, participants studied words either once (1x items) or three times (3x items), and during retrieval they were rewarded more either for detecting studied words (incentivize-old runs) or for detecting new words (incentivize-new runs). Regarding study repetitions, the attention to memory model assumes that VPC mediates bottom-up attention driven by salient memories, and hence it predicts greater VPC activity while detecting 3x than 1x items, as study repetition typically results in higher hit rates, shorter recognition times (e.g., ; ), and increased recognition confidence (). In contrast, the model assumes that DPC mediates top-down attention required by demanding search and monitoring processes, and therefore it predicts greater DPC activity for 1x than 3x items. Regarding retrieval goals, the model assumes that DPC mediates top–down attention driven by retrieval goals, hence the strategic orienting of attention toward different classes of items (old, new) depending on payoffs. Thus, the model predicts greater DPC activity for detection of memory targets (i.e., events consistent with retrieval goals, because rewarded more: old items in incentivize-old runs and new items in incentivize-new runs) than non-targets (i.e., new items in incentivize-old runs and old items in incentivize-new runs). Conversely, the model assumes that VPC mediates the bottom-up capture of attention by salient events inconsistent with retrieval goals, and therefore predicts greater VPC activity for detection of non-targets than of memory targets, in line with previous evidence of VPC involvement in invalidly cued and involuntary memory retrieval (reviewed in ).
Although our main predictions pertain to posterior parietal cortex, highly complex cognitive processes, such as episodic memory retrieval, are expectedly mediated by the interaction among functionally related regions. Therefore, we adopted a multivariate method, Partial Least Squares (PLS) (), to reveal the coordinated activity of distributed networks, supposedly including DPC and VPC, associated with top–down and bottom up attention to memory, respectively. In a previous work using the PLS method (), for example, we showed that DPC was functionally connected with a dorsal network of brain regions during cued (top–down) recognition memory trials (e.g., dorsolateral prefrontal cortex, precuneus), whereas VPC was functionally connected with a ventral network of brain regions during uncued (bottom–up) memory trials (e.g., ventrolateral prefrontal cortex, insula; ). One important question pertains to the interaction between posterior parietal cortex (DPC and VPC) and the hippocampus. The hippocampus is thought to act as an index to neocortical structures representing the perceptual, conceptual, and emotional details of complex events (Teyler and Rudy, 2007; Stella et al., 2012). Recent research, however, indicates that differences exist in the type of information represented by the anterior and posterior hippocampi, based on their connectivity (). The posterior hippocampus is preferentially connected to perceptual regions in the posterior neocortex, supporting fine-grained, perceptually based memory representations, whereas the anterior hippocampus is preferentially connected to anterior regions, such as the ventromedial prefrontal cortex (vmPFC) and the amygdala, supporting memory representations that are more abstract (schematic) and subject to the influence of emotional/motivational processes (see , for a review). Several studies have found evidence of connectivity between the posterior hippocampus and VPC (Uddin et al., 2010; ; ; ), and we found evidence of anterior hippocampus-DPC connectivity in a study examining top–down attention to memory (). Based on this preliminary evidence, we predicted that the anterior hippocampus, which is connected with regions involved in motivation and reward processing, would be functionally coupled with DPC, responding more to either old or new items depending on which was rewarded more. In contrast, the posterior hippocampus, which is involved in the recollection of detailed memories (; ), should be functionally coupled with VPC, and signal salient memories regardless of payoffs.
Materials and Methods
Participants
Fifteen young adults participated in the study (10 females), but a male subject was excluded due to reported discomfort in the scanner and consequently poor memory performance (d’ = −0.01). The final sample, therefore, comprised 14 young adults (age range 22–33, mean age 26 years), and the low sample size is one caveat of our study. For one participant, data from 2 out of the 8 recognition runs were lost due to a technical problem, and therefore data for this participant are relative to the remaining 6 runs. All participants were healthy, right-handed, English speakers, and with no psychiatric or neurological history. Participant received $60 to participate in the study, and an additional bonus of $30 depending on performance (see below).
Stimuli and Procedure
Five hundred and twelve words (mean frequency = 25.49, SD = 34.9; mean concreteness = 4.92, SD = 1.79), between 4 and 13 letters long, were selected from the pool. Half of the words referred to concrete entities (e.g., volcano; mean concreteness = 6.56) and the other half referred to abstract entities (e.g., democracy; mean concreteness = 3.28). The words were subdivided into 4 lists of 64 concrete words (matched for frequency and concreteness; p > 0.69 in both analyses) and 4 lists of 64 abstract words (matched for frequency and concreteness; p > 0.98 in both analyses), which were randomly attributed to the different experimental conditions, with the study status (studied, unstudied) counterbalanced across participants.
The experimental paradigm was composed of a study phase and a test phase. During the study phase, outside the scanner, participants made concrete/abstract judgments on 256 words (128 abstract + 128 concrete), of which half were presented once (1x items) and the other half were presented three times (3x items). Each word was presented for 2800 ms followed by a fixation cross, which was presented for 200 ms. A scanned recognition memory test followed immediately afterward, which consisted of 8 runs. In each run, participants were presented with 64 words: 16 words studied once (8 abstract + 8 concrete), 16 words studied 3 times (8 abstract + 8 concrete), and 32 new words (16 abstract + 16 concrete), and had to recognize them as studied or new by pressing one of two keys (counterbalanced), located on an MRI-compatible response pad. The beginning of each recognition trial was signaled by a fixation cross that stayed in the center of the screen for 500 ms, and the target word was then presented for 3000 ms. Each recognition run also comprised 4 null events, in which a meaningless stimulus (i.e., xxxxxxx) appeared on the screen in the place of the word. Subjects were instructed to look at it and press one of the two response keys. An inter-trial-interval (ITI) (without fixation cross) that varied randomly between 2000 and 6000 ms was interspersed across test trials to “jitter” the onset times of trials and allow for event-related fMRI analyses.
In order to modify retrieval goals, and hence “memory targetness”, we used a payoff manipulation (; ). In half of the runs, subjects were informed that they would be rewarded 5 points for each correct “old” response and 1 point for each correct “new” response (i.e., incentivize-old runs). Conversely, in the other half, subjects were informed that they would be rewarded 5 points for each correct “new” response and 1 point for each correct “old” response (i.e., incentivize-new runs). Our assumption is that the manipulation would orient participants’ (top–down) attention toward different classes of items (old vs. new) depending on payoffs, with old words being the target for memory search in incentivize-old runs, and new words in incentivize-new runs. After each test run, the subject’s score for that run was displayed. Subjects were told that the participant with the highest final score would be rewarded an extra $30 after the experiment was completed, and $30 were accordingly awarded to the highest scoring subject. The order of incentivize-old and incentivize-new runs and the assignment of test words to the different runs were randomized for each participant.
fMRI Data Acquisition and Pre-processing
Images were acquired at Baycrest Hospital on a 3 Tesla Siemens Magnetom Trio whole-body scanner with a matrix 12-channel head coil. Anatomical images were acquired using a MP-RAGE sequence (TR: 2 s, TE: 2.63 s, 160 oblique axial slices, with a 1 mm3 voxel size, FOV = 25.6 cm, acquisition matrix: 256 × 256). Brain activation was assessed using the blood oxygenation level dependent (BOLD) effect with optimal contrast (). Functional images were obtained using a whole head T2∗-weighted echo-planar image (EPI) sequence (repetition time, TR: 2 s, echo time, TE: 30 ms, flip angle: 70°, 28 oblique axial slices with interleaved acquisition, 3.1 × 3.1 × 5 mm voxel resolution, field of view, FOV: 20 cm, acquisition matrix: 64 × 64). Physiological data (heart and respiration rate) were acquired during the scanning session.
The fMRI data were preprocessed using the Analysis of Functional NeuroImages software (AFNI; ). The initial five time points from each image volume were removed from analyses to allow for the brain magnetization to stabilize. EPI time-series data were corrected for cardiac and respiratory parameters (program 3dretroicor). Time-series data were spatially co-registered (program 3dvolreg) to correct for small head motion, using a 3-D Fourier transform interpolation. Each run was then normalized based on the mean intensity of the signal. Individual analysis was performed by generating the hemodynamic response function model for each condition, based on the convolution of the time points beginning with the stimulus presentation, using a block function (). For each subject, 6 trial types of interest were modeled: (1) incentivize-old 3x hits, (2) incentivize-old 1x hits, (3) incentivize-old CRs, (4) incentivize-new 3x hits, (5) incentivize-new 1x hits, and (6) incentivize-new CRs. They were modeled by fitting a general linear model to the measured fMRI time series at each voxel (program 3dDeconvolve). The number of trials was > 22 in each of the 6 conditions of interest, for all subjects. Null events, false alarms, and misses were also modeled but were not used in the analyses. Prior to group analyses, the activation maps for each participant and each condition were spatially normalized to an average volume of 152 normal skull stripped brains. Datasets were then re-sampled with a 2 × 2 × 2 voxel dimension (program @auto_tlrc) and spatially smoothed with a 8 mm full-width half-maximum Gaussian kernel (program 3dmerge).
fMRI Data Analysis
Whole-Brain Analysis
As memory processing is the result of integrated and coordinated activity of groups of brain regions (i.e., distributed brain networks) rather than the independent activity of any single brain region, fMRI data were analyzed with the Partial Least Squares multivariate analytical technique (PLS; ; ; ; for a detailed tutorial and review of PLS, see ), which is designed to identify groups of brain regions distributed over the entire brain whose activity changes as a function of task demands or is correlated with behavioral performance. PLS uses singular value decomposition (SVD) of a single matrix that contains all participants’ data to find a set of latent variables (LVs), which are mutually orthogonal dimensions that reduce the complexity of the data set. In the current study, we used whole-brain PLS to examine changes in activity in the six experimental conditions of interest. The output of PLS analysis is a set of LVs reflecting cohesive patterns of brain activity related to the experimental design, and accounting for maximum covariance between regional activity changes and task conditions. Thus, akin to Principal Component Analysis (PCA; e.g., ), PLS enables us to differentiate the degree of contribution of different brain regions associated with task or performance. Each LV consists of a singular image of voxel saliences (i.e., a spatiotemporal pattern of brain activity that reflects task-related changes or brain-behavior correlations seen across conditions), a singular profile of task saliences (i.e., a set of weights that indicate how brain activity in the singular image is expressed in each of the experimental conditions), and a singular value (i.e., the amount of covariance accounted for by the LV). The first LV always accounts for the largest amount of covariance (i.e., has the largest singular value), with subsequent LVs accounting for progressively smaller amounts. For each condition in each LV, we calculated summary measures of how strongly each participant expresses the particular pattern of activity seen on the LV. These measures, called brain scores, are the products of the weighted salience of each voxel and BOLD signals summed across the entire brain for each participant in each condition on a given LV (). Salience indicates the degree to which a voxel is related to the LV and can be positive or negative, depending on the relation of the voxel to the pattern of task-dependent differences identified by the LV. The significance and reliability of each LV was determined by permutation and bootstrap resampling tests (see below).
Functional Connectivity Analysis
In addition to whole-brain PLS analysis, we examined task-related functional connectivity (i.e., the degree of non-zero correlation between brain regions), using the ‘seed’ PLS analysis (; Schreurs et al., 1997). Seed PLS is a multivariate statistical method widely used to investigate the relation between activity in a selected brain region (seed voxel) and activity in the rest of the brain, across task conditions (; Schreurs et al., 1997; ; ). Based on previous evidence on the differential role of the anterior and posterior hippocampus in episodic memory retrieval (), and on the findings from the whole-brain PLS analysis, functional seed values were extracted from a region of interest with a neighborhood size of one voxel (i.e., including the seed voxel plus one voxel adjacent to the peak voxel in each direction; see also ; Ziaei et al., 2017; ) centered in the left anterior hippocampus (MNI coordinates: x = −30, y = −10, z = −20) and the left posterior hippocampus (MNI coordinates: x = −26, y = −26, z = −22), to examine, respectively, task-related functional connectivity during detection of memory targets (i.e., including hits in the incentive-old condition and CRs in the incentive-new condition in the analysis) and detection of items that were not the target of memory (i.e., including hits in the incentive-new condition and CRs in the incentive-old condition in the analysis). The analytical procedure for the seed PLS functional connectivity analysis was the following: first, the BOLD values from the hippocampal seed regions were extracted for each event of interest (detection of memory targets and non-targets) across 8 time points from the onset of the trial. The activity for each seed region was averaged across the peak and adjacent timepoints, and then this average measure of seed activity was correlated with activity in all other brain voxels, across participants, within each condition of interest. These correlations were then combined into a matrix and decomposed with singular value decomposition (SVD), resulting in a set of LVs and voxel saliences.
The significance level for each LV is tested via two steps: permutations and bootstrap estimation, which is the standard analytical approach in PLS (e.g., ; ; Vallesi et al., 2009; St-Laurent et al., 2011; , ; Ziaei et al., 2017; ). The significance for each LV as a whole is determined using a permutation test (). The permutation test samples the distribution by resampling the observed data, testing the hypothesis of whether the whole-brain activity during a task/condition significantly differs from noise. At each permutation, the data matrix rows are randomly reordered and a new set of LVs is calculated each time. The singular value of each new LV is compared to the singular value of the original LV. A probability is assigned to the initial value based on the number of times a statistic from the permuted data exceeds this original value (). For the current experiment, 500 permutations were used. If the probability was less than 0.05 then the LV was considered significant. This first step is then followed by a bootstrap test providing a direct assessment of the reliability of the significant patterns identified by the permutation test, and allows estimating voxel saliences, which are weights indicating how strongly a given voxel contributes to a significant LV. To determine the reliability of the saliences for the voxels characterizing each pattern identified by the LVs, all data were submitted to a bootstrap estimation of the standard errors, by randomly re-sampling subjects with replacement 100 times. PLS is recalculated for each bootstrap sample to identify those saliences whose value remains stable regardless of the sample chosen (Sampson et al., 1989). The ratio of the salience to the bootstrap standard error (bootstrap ratio, BSR) is approximately equivalent to a z score given a normal bootstrap distribution (). Peak voxels with a BSR > 3 (approximately equivalent to a z-score corresponding to p < 0.001) were considered as reliable. Since in PLS multivariate methods the whole-brain spatiotemporal patterns are derived in a single analytical step (via SVD, ), there is no need for correction for multiple comparisons.
Results
Behavior
Behavioral results are summarized in Table 1 and Figure 1. An analysis of variance (ANOVA) on the frequency of correct responses with Item (1x, 3x, new) and Run (incentivize-old, incentivize-new) as within-subject factors yielded an effect of Item, F(2,26) = 38.10, p < 0.001, ηp2 = 0.75, qualified by a significant Item X Run interaction, F(2,26) = 19.28, p < 0.001, ηp2 = 0.60. Post hoc comparisons, performed with the Scheffè test, indicated that, as expected, hit rates for 3x items were higher than hit rates for 1x items (Figure 1A) in both the incentivize-old and incentivize-new condition (p < 0.001 in both cases). This result confirms the effectiveness of our study repetition manipulation. Importantly, hit rates for 1x items were higher in the incentivize-old compared to the incentivize-new condition (p = 0.02), whereas CR rates were higher in the incentivize-new than in the incentivize-old condition (p = 0.01). This result suggests that participants changed their retrieval orientation depending on whether hits or CRs were rewarded more, confirming the effectiveness of our targetness manipulation (Figure 1B). By contrast, hit rates for 3x items were comparable between incentivize-old and incentivize-new runs (p = 0.39). Arguably, the fact that participants had a very high recognition performance with 3x items rendered their ‘old’ status more obvious, and their recognition less sensitive to the payoff manipulation compared to 1x items (Figure 1B).
TABLE 1
| Hit Rates | CR Rates | Sensitivity (d’) | Criterion (C) | RTs (correct responses) | ||||
| 1x | 3x | New | 1x | 3x | New | |||
| Incentivize-old condition | 0.74 (0.03) | 0.93 (0.02) | 0.71 (0.03) | 1.62 (0.14) | −0.22 (0.07) | 1271 (54) | 1089 (44) | 1424 (67) |
| Incentivize-new condition | 0.63 (0.02) | 0.86 (0.02) | 0.83 (0.03) | 1.72 (0.17) | 0.16 (0.05) | 1383 (62) | 1165 (48) | 1371 (56) |
Behavioral data.
1x, items studied once; 3x, items studied 3 times; CR, correct rejection; RTs, response times. The values in parenthesis are standard errors of the mean.
FIGURE 1
A similar ANOVA on response times (RTs) for correct responses showed an effect of Item, F(2,26) = 41.60, p < 0.001, ηp2 = 0.76, qualified by a significant Item X Run interaction, F(2,26) = 7.07, p = 0.003, ηp2 = 0.35. Post hoc Scheffè comparisons showed that individuals were faster at recognizing 3x than 1x items (Figure 1A), in both the incentivize-old and incentivize-new conditions (p < 0.001 in both cases), again confirming the efficacy of our study repetition manipulation (Figure 1A). RTs for correctly recognizing 1x items (p = 0.07), 3x items (p = 0.39), and new items (p = 0.74) did not change significantly between the incentivize-old and incentivize-new conditions, although participants tended to be faster at recognizing 1x items in the incentivize-old condition (when they were the target of memory search; see Table 1).
We also report the estimates of response bias and sensitivity (collapsing hit rates for 1x items and 3x items; see Table 1). Response bias was estimated with a criterion location measure, defined as c = 0.5[z(H) + z(F)] (). Negative c values indicate a liberal response bias, whereas positive values indicate a conservative response bias, and expectedly participants exhibited lower c values in the incentivize-old than in the incentivize-new condition, t(13) = 4.86, p < 0.001. In contrast, sensitivity, estimated as d = z(H) – z(F) (), did not differ significantly between conditions, t(13) = 1.24, p = 0.23. A two-one-sided test for equivalence (TOST; ), however, indicated that the observed effect size (Cohen’s dz = −0.26) was not significantly within the equivalence bounds of dz = −0.50 and dz = 0.50, and therefore it was not statistically consistent with a lack of a medium effect-size result, t(13) = 0.89; p = 0.196, though it could exclude a large effect-size result, t(13) = 2.01; p = 0.033.
fMRI
Whole-Brain Analysis
We report patterns of brain activity related to study repetitions (including 3x hits vs. 1x hits) and retrieval goals (including memory targets, i.e., hits in the incentive-old condition and CRs in the incentive-new condition, vs. non-targets, i.e., hits in the incentive-new condition and CRs in the incentive-old condition). In the targetness analysis, we included only 1x items, because the mnemonic status of these items is less obvious and more influenced by criterion manipulations than that of 3x items, consistent with the results obtained on hit rates (see Table 1 and Figure 1B; see also Stretch and Wixted, 1998; , for similar findings). Including all hits led to a similar pattern of results.
Study repetitions
The statistically significant LV (p = 0.038) delineated a whole pattern of brain regions that responded differentially to 3x and 1x hits (Figures 2A,B). In line with previous research (e.g., ; ; ; ), detection of items studied three times vs. once was associated with activity in VPC (supramarginal gyrus; Figure 2A) bilaterally (p < 0.001; see Table 2 for the complete list of activations). Consistent with our hypotheses, the 3x study repetitions pattern also included the right posterior hippocampus. Activity was also detected in the parahippocampal gyri and a network of brain regions including the anterior and lateral prefrontal cortex. In contrast, in line with previous evidence (e.g., ; ; ), detecting items studied only once was associated with activity in the left DPC (superior parietal lobule), although activity in the angular gyrus bilaterally was also detected (Figure 2B). This pattern of brain activity also included a more anterior region of the left hippocampus, as well as the medial and lateral prefrontal cortex and the anterior cingulate cortex (see Table 2).
FIGURE 2
TABLE 2
| Area | BA | X | Y | Z | BSR |
| 3x HITs > 1x HITs | |||||
| Ventral parietal cortex (supramarginal gyrus) | 40 | −56 | −44 | 50 | 6.01 |
| 40 | 58 | −36 | 46 | 3.01 | |
| Posterior hippocampus | 20 | −36 | 8 | 3.28 | |
| Parahippocampal gyrus | 36 | −44 | −32 | −16 | 4.92 |
| 35 | 32 | −22 | −26 | 5.47 | |
| Frontopolar cortex | 10 | 34 | 46 | 12 | 5.26 |
| Dorsolateral prefrontal cortex | 46 | −40 | 38 | 8 | 4.08 |
| 9 | 48 | −2 | 18 | 4.73 | |
| Lateral temporal cortex | 21 | −66 | −38 | −8 | 4.18 |
| 21 | 58 | −22 | −12 | 3.77 | |
| 41 | 50 | −36 | 4 | 4.75 | |
| Insula | 13 | −36 | 6 | 16 | 3.46 |
| 13 | 50 | 6 | −2 | 3.55 | |
| 1x HITs > 3x HITs | |||||
| Dorsal parietal cortex | 7 | −12 | −52 | 68 | 4.23 |
| Ventral parietal cortex (angular gyrus) | 39 | −36 | −70 | 30 | 4.40 |
| 39 | 36 | −64 | 36 | 5.73 | |
| Middle/posterior Hippocampus | −37 | −24 | −12 | 4.42 | |
| Parahippocampal gyrus | 36 | 38 | −28 | −18 | 6.20 |
| Medial prefrontal cortex | 9 | −6 | 52 | 14 | 4.74 |
| 10 | −2 | 56 | 0 | 3.48 | |
| 10 | −28 | 50 | 6 | 3.41 | |
| Dorsolateral prefrontal cortex | 9 | 40 | 22 | 32 | 3.22 |
| Ventrolateral prefrontal cortex | 47 | −38 | 38 | −6 | 4.27 |
| Anterior cingulate cortex | 24 | −24 | −6 | 36 | 3.28 |
| 24 | 6 | −16 | 38 | 4.37 | |
| Posterior cingulate cortex | 31 | −22 | −64 | 18 | 4.27 |
| 23 | 4 | −34 | 22 | 5.85 | |
| Lateral temporal cortex | 21 | −56 | −12 | −20 | 4.92 |
| 22 | −46 | −22 | 4 | 5.47 | |
| 38 | 42 | 2 | −22 | 5.71 | |
Networks related to study repetitions.
Key: 1x HITs, hits to items studied once; 3x HITs, hits to items studied three times; BA, Brodmann area; BSR, salience/standard error ratio from the bootstrap analysis. We report Montreal Neurological Institute (MNI) coordinates.
Targetness
The statistically significant LV (p = 0.046) delineated a whole pattern of brain regions that responded differentially to detection of memory targets and non-targets (Figures 3A,B). Consistent with our hypotheses, detection of memory targets was associated with activity in the left DPC (superior parietal lobule and precuneus; Figure 3A), along with a relatively dorsal region of the inferior parietal lobule bilaterally (p < 0.001; see Table 3 for the complete list of activations). The ‘targetness pattern’ also included the anterior hippocampus, and the ventrolateral prefrontal and anterior cingulate cortex. A different pattern of brain regions evinced higher activation for items that were not in line with retrieval goals. Consistent with our hypotheses, these included prominently VPC, with multiple peaks in the inferior parietal lobule and supramarginal gyrus, in addition to the precuneus (Figure 3B). The left posterior hippocampus was also activated, along with a more anterior region of the right hippocampus, and the medial and dorsolateral prefrontal cortex (see Table 3).
FIGURE 3
TABLE 3
| Area | BA | x | y | z | BSR |
| Memory targets > Non-targets | |||||
| Dorsal parietal cortex | 7 | −36 | −50 | 64 | 4.30 |
| Precuneus | 7 | −12 | −54 | 50 | 5.39 |
| Ventral parietal cortex (inferior parietal lobule) | 40 | −52 | −48 | 48 | 4.48 |
| 40 | 36 | −46 | 48 | 4.47 | |
| Anterior hippocampus | −30 | −10 | −20 | 3.95 | |
| Ventrolateral prefrontal cortex | 47 | −38 | 50 | −10 | 4.44 |
| Anterior cingulate cortex | 32 | −4 | 30 | 34 | 5.82 |
| Lateral temporal cortex | 21 | −62 | −42 | −12 | 4.61 |
| Temporo-occipital cortex | 37 | −38 | −58 | −12 | 6.26 |
| Occipital cortex | 19 | −28 | −70 | 6 | 5.09 |
| 19 | 24 | −82 | 6 | 6.60 | |
| Insula | 13 | −26 | 10 | 12 | 4.56 |
| Non-targets > memory targets | |||||
| Ventral parietal cortex (supramarginal gyrus) | 40 | −62 | −40 | 24 | 5.56 |
| 40 | 56 | −42 | 24 | 4.75 | |
| 40 | −64 | −42 | 36 | 6.05 | |
| 40 | 64 | −40 | 36 | 5.61 | |
| Precuneus | 7 | 6 | −60 | 32 | 4.66 |
| Posterior hippocampus | −26 | −26 | −22 | 5.49 | |
| Anterior/middle hippocampus | 20 | −20 | −20 | 3.35 | |
| Medial prefrontal cortex | 10 | 10 | 48 | −6 | 4.70 |
| Dorsolateral prefrontal cortex | 46 | −30 | 46 | 16 | 8.74 |
| 46 | 44 | 46 | 10 | 4.42 | |
| Lateral temporal cortex | 21 | −48 | −38 | 4 | 6.06 |
| 21 | 48 | −8 | −10 | 5.56 | |
| Temporo-occipital cortex | 37 | −54 | −64 | 6 | 4.86 |
| 37 | 58 | −66 | 12 | 9.64 | |
Networks related to memory targetness.
Key: Memory targets, hits in the Incentive-old condition and CRs in the Incentive-new condition; Non-targets, hits in the Incentive-new condition and CRs in the Incentive-old condition; BA, Brodmann area; BSR, salience/standard error ratio from the bootstrap analysis. We report Montreal Neurological Institute (MNI) coordinates.
Functional Connectivity of the Hippocampus During Detection of Memory Targets and Non-targets
To test our hypothesis that the left anterior hippocampus would be functionally connected to DPC during detection of memory targets, and the left posterior hippocampus would be functionally connected to VPC during detection of non-targets, we investigated their task-related functional connectivity during detection of memory targets (hits in the liberal condition and CRs in the conservative condition) and non-targets (hits in the conservative condition and CRs in the liberal condition). The results confirmed that the left anterior hippocampus was functionally connected to the DPC during detection of memory targets but not during detection of non-targets (p < 0.001; see Table 4 for the complete list of activations). Regions functionally connected to the anterior hippocampus for target detection also included VPC (supramarginal gyrus), the anterior cingulate cortex, and the ventrolateral prefrontal cortex. These functional associations were significant for hits in the incentive-old condition, but not for CRs in the incentive-new condition. By comparison, the left posterior hippocampus was functionally connected to VPC (inferior parietal lobule and supramarginal gyrus) bilaterally during detection of items that were not the target for memory search (including both hits in the incentive-new condition and CRs in the incentive-old condition) but not during detection of memory targets (p < 0.001; see Table 4 for the complete list of activations). Regions functionally connected to the posterior hippocampus also included the dorsolateral prefrontal cortex and the right anterior hippocampus. The supramarginal gyrus of VPC exhibited a significant positive correlation with the left posterior hippocampus for CRs in the incentive-old condition, but a negative correlation for hits in the incentive-new condition, as did the dorsolateral prefrontal cortex, suggesting a selective engagement in signaling unexpected novelty (see also ). Because detection of non-targets also entailed activity in a right anterior hippocampal region, for completeness we ran the same functional connectivity analysis using this region as the seed (Table 3). The anterior hippocampus seed was functionally connected to the left posterior hippocampus (p < 0.001), but less connected to the anti-targetness network itself: for hits in the incentive-new condition, but not for CRs in the incentive-old condition, it was functionally associated with the right supramarginal gyrus, the inferior parietal lobule, and the left dorsolateral prefrontal cortex, but showed a negative correlation with the left supramarginal gyrus and the right dorsolateral prefrontal cortex.
TABLE 4
| Area | BA | x | y | z | BSR |
| Detection of memory targets | |||||
| Anterior hippocampus (seed) | −30 | −10 | −20 | 3.95 | |
| Dorsal parietal cortex | 7 | −36 | −50 | 64 | 4.30 |
| Ventral parietal cortex | 40 | 36 | −46 | 48 | 4.47 |
| Anterior cingulate cortex | 32 | −4 | 30 | 34 | 5.82 |
| Ventrolateral prefrontal cortex | 47 | −38 | 50 | −10 | 4.44 |
| Detection of non-targets | |||||
| Posterior hippocampus (seed) | −26 | −26 | −22 | 5.49 | |
| Anterior/middle hippocampus | 34 | −20 | −20 | 3.35 | |
| Ventral parietal cortex | 40 | −62 | −40 | 24 | 5.56 |
| 40 | 56 | −42 | 24 | 4.75 | |
| 40 | −64 | −42 | 36 | 6.05 | |
| Dorsolateral prefrontal cortex | 46 | −30 | 46 | 16 | 8.74 |
| 46 | 44 | 46 | 10 | 4.42 | |
Hippocampal functional connectivity.
BA, Brodmann area; BSR, salience/standard error ratio from the bootstrap analysis. We report Montreal Neurological Institute (MNI) coordinates.
Discussion
The first goal of the present study was to test the attention to memory model of posterior parietal contributions to episodic memory retrieval during a standard recognition memory task by manipulating study repetitions, and supposedly the saliency of recovered memories at retrieval, and memory goals, hence memory targetness. The second goal of the study was to investigate the functional connectivity of posterior parietal regions during bottom–up and top–down attention to memory, and, in particular, the interaction between posterior parietal regions and the posterior and anterior hippocampus. We had three main predictions. First, we predicted that VPC would be more active for detection of 3x than 1x items, whereas DPC would be more active for 1x than 3x items. Second, we predicted that DPC would be more active for detection of memory targets than non-targets, whereas VPC would be more active for non-targets than targets. Finally, we predicted that VPC connectivity would be stronger with the posterior hippocampus, whereas DPC connectivity would be stronger with the anterior hippocampus. The results were generally consistent with our predictions, but there were also some unpredicted findings.
Effect of Study Repetitions
Consistent with attention to memory model (; ), correct recognition of 3x items (vs. 1x) was associated with increased activity in the supramarginal gyrus of VPC, consistent with the hypothesis that VPC signals retrieval of salient memories that capture attention in a bottom–up fashion (), and not in DPC. Items that have been studied multiple times are indeed generally recognized quickly and with high confidence (; ), which is also associated with the engagement of VPC (; ; ; ). Also consistent with the attention to memory model, hits for 1x (vs. 3x) items were associated with increased DPC activity: 1x items likely passed through more pre- and post-retrieval processing before being endorsed as old, requiring the sustained deployment of attentional resources (; Sestieri et al., 2010; ).
One unexpected result was the finding of greater activity for 1x than 3x hits in the angular gyrus within VPC. An influential hypothesis maintains that the angular gyrus acts as an episodic buffer to hold integrated representations retrieved from episodic memory in the service of memory decisions (; Vilberg and Rugg, 2007; ; Shimamura, 2011; see also ; ). Activity in VPC, indeed, has been found to increase with the amount of information recollected (Vilberg and Rugg, 2007, 2009). It is possible, therefore, that the less obvious mnemonic status of 1x than 3x items made them more behaviorally relevant, because more susceptible to the payoff manipulation, as borne out in the behavioral data (Figure 1). Therefore, activity in the angular gyrus may reflect the prolonged online maintenance of 1x memories needed to integrate memory signals with payoffs in order to drive adaptive decisions and earn points, as suggested by increased RTs (Table 1). This interpretation is compatible with the view that the angular gyrus supports an episodic buffer for retrieved information in the service of memory decisions, whereas the supramarginal gyrus mediates effects more directly related to bottom-up attention (; Sestieri et al., 2017).
Effect of Targetness
Consistent with the attention to memory model (; ), DPC (superior parietal lobe and precuneus) was sensitive to memory targetness, responding strongly to hits in the incentive-old condition and CRs in the incentive-new condition, consistent with a role in top-down attention to memory. The successful retrieval of memory targets was also marked by activity in a dorsal, anterior region of the inferior parietal lobe, along with nodes of the salience network (Seeley et al., 2007), such as the anterior cingulate cortex and the insula.
This finding makes contact with previous studies showing that activity in the posterior parietal cortex is related to response bias (; ; ). A common finding of these studies is that DPC regions respond more to recognition hits when studied items are infrequent (vs. frequent), which typically results in a more conservative criterion (; Vilberg and Rugg, 2009; ; ). Note that the task used in most previous studies required detecting studied words (memory targets; ), and when studied items are few, their behavioral relevance increases further. DPC may thus index the behavioral relevance (targetness) of retrieved items (; ), be this determined by mnemonic expectations (Vilberg and Rugg, 2009; ; ; see also ) or payoffs (this study). In particular, we argue that DPC mediated the top-down orienting of attention toward different classes of items (old, new) depending on payoffs, consistent with reduced RTs for memory targets compared to non-targets (as is observed for valid trials in the Posner task). Our results are also consistent with previous evidence that posterior parietal cortex tracks retrieval goals, even though this is not consistently confined to the DPC (e.g., ; ). For example, found that DPC -but not VPC- represented more goal-relevant than goal-irrelevant feature information at retrieval. found that DPC and the supramarginal gyrus (but not the angular gyrus) were more active during goal-relevant vs. incidental reactivation of event features. found that a VPC region responded more to hits when incentives were paired with old compared to new recognition memory decisions. With respect to the striatum, however, our findings diverged from those of ; for review see Scimeca and Badre, 2012, and references therein), in that we did not find differential striatal responses depending on targetness. This discrepancy may be related to the fact that our design manipulated the amount of reward associated with hits and CRs (1 vs. 5 points), but did not contain a condition with no reward, or with punishment (unlike ).
A different neural network was engaged when participants detected items that were not the target of memory search, because not in line with retrieval goals. In this case, we observed bilateral activity in multiple sites of VPC, including the supramarginal gyrus and a ventral region of the inferior parietal lobule, and no activity in DPC. This finding is consistent with the hypothesis that VPC signals the bottom–up reorienting of attention to salient yet unattended memories, and aligns with previous evidence that VPC activity is associated with unintentional memory retrieval (; ; ), with retrieval of items that were invalidly (vs. validly) cued (; ; ), with retrieval of items overcoming active suppression (), and even with mind-wandering, the automatic drift of attention away from an external task toward inner thoughts (e.g., memories; ), which consistently engages VPC but not DPC ().
Parietal-Hippocampal Connectivity
The results show that the left anterior hippocampus was associated with the detection of memory targets, along with DPC, whereas the left posterior hippocampus was associated with the detection of targets not aligned with memory goals, along with VPC. The response profiles of the anterior and posterior hippocampi are consistent with the recently described functional organization along the hippocampal antero-posterior axis, according to which the anterior hippocampus supports coarse memory representations, subject to the influence of schematic knowledge and motivational factors, whereas the posterior hippocampus supports fine-grained representations related to recollection abilities (; ; Strange et al., 2014; ; ; ). In our study, indeed, the anterior hippocampus was influenced by retrieval goals rather than its objective memory status, whereas the posterior hippocampus supported memory decisions not influenced by (in fact, in conflict with) retrieval goals. Other studies have found that motivational salience modulates activity in the anterior hippocampus (; Zweynert et al., 2011). In other words, the ‘stupidity’ quality provocatively attributed to the hippocampus by to describe the obligatory nature of retrieval () appears to apply to its posterior sector only, as the anterior hippocampus can be made to care about mnemonic goals and reward.
Other regions functionally connected to the anterior hippocampus in signaling targetness were a dorsal region of the anterior cingulate cortex, which has been associated with cognitive control (; ; ), and the ventrolateral prefrontal cortex, which has been linked with the selection of task-relevant memory contents (). These regions were likely necessary to monitor participants’ retrieval goals along with the objective memory status of items, in order to favor rewarding response strategies. The anterior hippocampus was also functionally connected with a region in the right inferior parietal lobe, possibly mediating detection of task-relevant memory contents. The functional connectivity of the left posterior hippocampus involved, in addition to VPC regions, a more anterior region of the right hippocampus, perhaps encoding the ‘contextual novelty’ of items violating mnemonic expectations (; ; ), and a dorsal prefrontal region widely implicated in post-retrieval evaluation of memory output with respect to task relevance and accuracy (; ; ).
In conclusion, our results show that DPC, within a network of brain regions functionally connected to the anterior hippocampus, is associated with top–down attention to memory retrieval, supporting retrieval of items consistent with memory goals and of items of uncertain memory status. By comparison, VPC, within a network of brain regions functionally connected to the posterior hippocampus, is more prominently associated with retrieval of salient memories, and of retrieval cues not aligned with the current goals and mental sets, which both capture attention bottom–up.
Limitations
A caveat of this study is the small sample size. It would be important, therefore, to confirm our finding of a differential role of DPC and VPC in top–down and bottom–up attention to memory with larger samples, or to seek complementary, causative evidence for this dissociation, for example testing patients with focal lesions to DPC or VPC or interfering with these regions with transcranial magnetic stimulation (TMS).
In addition, although our results are generally consistent with a dorsal/ventral functional partition of posterior parietal cortex during episodic memory retrieval, we have also found evidence, as other have (, ; Sestieri et al., 2017), that VPC does not behave as a single functional unit. In the present study, for example, the supramarginal and the angular gyrus responded preferentially during detection of items studied multiple times and once, respectively. Thus, overarching single-function accounts of VPC, such as the attention to memory model (), will need to be modified, or supplemented, to take such findings into account. On the other hand, we note that a similar functional heterogeneity characterizes the attentional properties of VPC. For example, TMS evidence shows that the angular (but not the supramarginal) gyrus is critical for reorienting attention after invalid cueing (; ; see also ). Future studies will clarify whether posterior parietal cortex has multiple mnemonic properties or, rather, episodic memory retrieval engages multiple facets of attention.
Statements
Data availability statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Ethics statement
The studies involving human participants were reviewed and approved by Roman Research Institute. The patients/participants provided their written informed consent to participate in this study.
Author contributions
EC, MM, and RC conceived the study. EC, HB, and AV collected and analyzed the data. EC wrote the first draft of the manuscript. All the authors edited the manuscript and approved its final version.
Funding
This work was supported by a Marie Curie Outgoing fellowship (#OIF-40575) to EC and an NSERC Grant (# A 8347) to MM.
Acknowledgments
We thank Buddhika Bellana, Marshall Dalton, and Conny McCormick for their comments on the manuscript.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
References
1
AminoffE. M.ClewettD.FreemanS.FrithsenA.TipperC.JohnsonA.et al (2012). Individual differences in shifting decision criterion: a recognition memory study.Mem. Cogn.401016–1030. 10.3758/s13421-012-0204-6
2
AminoffE. M.FreemanS.ClewettD.TipperC.FrithsenA.JohnsonA.et al (2015). Maintaining a cautious state of mind during a recognition test: a large-scale fMRI study.Neuropsychologia6132–147. 10.1016/j.neuropsychologia.2014.12.011
3
Andrews-HannaJ. R.ReidlerJ. S.SepulcreJ.PoulinR.BucknerR. L. (2010). Functional–anatomic fractionation of the brain’s default network.Neuron65550–562. 10.1016/j.neuron.2010.02.005
4
AssadJ. A. (2003). Neural coding of behavioral relevance in parietal cortex.Curr. Opin. Neurobiol.13194–197. 10.1016/s0959-4388(03)00045-x
5
BaddeleyA. (2003). Working memory: looking back and looking forward.Nat. Rev. Neurosci.4829–839. 10.1038/nrn1201
6
BadreD.WagnerA. D. (2007). Left ventrolateral prefrontal cortex and the cognitive control of memory.Neuropsychologia452883–2901. 10.1016/j.neuropsychologia.2007.06.015
7
BenoitR. G.AndersonM. C. (2012). Opposing mechanisms support the voluntary forgetting of unwanted memories.Neuron76450–460. 10.1016/j.neuron.2012.07.025
8
Ben-ZviS.SorokerN.LevyD. A. (2015). Parietal lesion effects on cued recall following pair associate learning.Neuropsychologia73176–194. 10.1016/j.neuropsychologia.2015.05.009
9
BerryhillM. E.PhuongL.PicassoL.CabezaR.OlsonI. R. (2007). Parietal lobe and episodic memory: bilateral damage causes impaired free recall of autobiographical memory.J. Neurosci.2714415–14144. 10.1523/jneurosci.4163-07.2007
10
BonniciH. M.MaguireE. A. (2018). Two years later - revisiting autobiographical memory representations in vmPFC and hippocampus.Neuropsychologia110159–169. 10.1016/j.neuropsychologia.2017.05.014
11
BonniciH. M.RichterF. R.YazarY.SimonsJ. S. (2016). Multimodal feature integration in the angular gyrus during episodic and semantic retrieval.J. Neurosci.365462–5471. 10.1523/jneurosci.4310-15.2016
12
BraverT. S.BarchD. M.GrayD. M.MolfeseD. L.SnyderA. (2001). Anterior cingulate cortex and response conflict: effects of frequency, inhibition and errors.Cereb. Cortex11825–836. 10.1093/cercor/11.9.825
13
BurianováH.CiaramelliE.GradyC. L.MoscovitchM. (2012). Top-down and bottom-up attention-to-memory: mapping functional connectivity in two distinct networks that underlie cued and uncued recognition memory.Neuroimage631343–1352. 10.1016/j.neuroimage.2012.07.057
14
BurianovaH.GradyC. L. (2007). Common and unique neural activations in autobiographical, episodic, and semantic retrieval.J. Cogn. Neurosci.191520–1534. 10.1162/jocn.2007.19.9.1520
15
BurianováH.MarstallerL.SowmanP.TesanG.RichA. N.WilliamsM.et al (2013). Multimodal functional imaging of motor imagery using a novel paradigm.Neuroimage7150–58. 10.1016/j.neuroimage.2013.01.001
16
CabezaR.CiaramelliE.MoscovitchM. (2012). Cognitive contributions of the ventral parietal cortex: an integrative account.Trends Cogn. Sci.16338–352. 10.1016/j.tics.2012.04.008
17
CabezaR.CiaramelliE.OlsonI. R.MoscovitchM. (2008). The parietal cortex and episodic memory: an attentional account.Nat. Rev. Neurosci.9613–625. 10.1038/nrn2459
18
CabezaR.MazuzY. S.StokesJ.KragelJ. E.WoldorffM. G.CiaramelliE.et al (2011). Overlapping parietal activity in memory and perception: evidence for the Attention to Memory (AtoM) model.J. Cogn. Neurosci.113209–3217. 10.1162/jocn_a_00065
19
CarterC. S.BraverT. S.BarchD. M.BotvinickM. M.NollD.CohenJ. D. (1998). Anterior cingulate cortex, error detection, and the online monitoring of performance.Science280747–749. 10.1126/science.280.5364.747
20
ChambersC. D.PayneJ. M.StokesM. G.MattingleyJ. B. (2004). Fast and slow parietal pathways mediate spatial attention.Nat. Neurosci.7217–218. 10.1038/nn1203
21
CiaramelliE.FaggiG.ScarpazzaC.MattioliF.SpaniolJ.GhettiS.et al (2017). Subjective recollection independent from multifeatural context retrieval following damage to the posterior parietal cortex.Cortex91114–125. 10.1016/j.cortex.2017.03.015
22
CiaramelliE.GradyC.LevineB.WeenJ.MoscovitchM. (2010a). Top-down and bottom-up attention-to-memory are dissociated in posterior parietal cortex: functional neuroimaging and neuropsychological evidence.J. Neurosci.304943–4956. 10.1523/jneurosci.1209-09.2010
23
CiaramelliE.RosenbaumR. S.SolczS.LevineB.MoscovitchM. (2010b). Mental space travel: damage to posterior parietal cortex prevents egocentric navigation and reexperiencing of remote spatial memories.J. Exp. Psychol. Learn. Mem. Cogn.36619–634. 10.1037/a0019181
24
CiaramelliE.GradyC. L.MoscovitchM. (2008). Top-down and bottom-up attention to memory: a hypothesis (AtoM) on the role of the posterior parietal cortex in memory retrieval.Neuropsychologia461828–1851. 10.1016/j.neuropsychologia.2008.03.022
25
CiaramelliE.MoscovitchM. (2020). The space for memory in the posterior parietal cortex.Neuropsychologia146:107551. 10.1016/j.neuropsychologia.2020.107551
26
CorbettB.RajahM. N.DuarteA. (2020). Preparing for the worst: evidence that older adults proactively downregulate negative affect.Cereb. Cortex301291–1306. 10.1093/cercor/bhz166
27
CorbettaM.KincadeJ. M.OllingerJ. M.McAvoyM. P.ShulmanG. L. (2000). Voluntary orienting is dissociated from target detection in human posterior parietal cortex.Nat. Neurosci.3292–297. 10.1038/73009
28
CorbettaM.PatelG.ShulmanG. (2008). The reorienting system of the human brain: from environment to theory of mind.Neuron58306–324. 10.1016/j.neuron.2008.04.017
29
CorbettaM.ShulmanG. L. (2002). Control of goal-directed and stimulus-driven attention in the brain.Nat. Rev. Neurosci.3201–215. 10.1038/nrn755
30
CoxR. W. (1996). AFNI: software for analysis and visualization of functional magnetic resonance neuroimages.Comput. Biomed. Res.29162–173. 10.1006/cbmr.1996.0014
31
DaselaarS. M.FleckM. S.CabezaR. (2006). Triple dissociation in the medial temporal lobes: recollection, familiarity, and novelty.J. Neurophysiol.961902–1911. 10.1152/jn.01029.2005
32
DaselaarS. M.HuijbersW.EklundK.MoscovitchM.CabezaR. (2013). Resting-state functional connectivity of ventral parietal regions associated with attention reorienting and episodic recollection.Front. Hum. Neurosci.22:38. 10.3389/fnhum.2013.00038
33
DavidsonP. S.AnakiD.CiaramelliE.CohnM.KimA. S.MurphyK. J.et al (2008). Does lateral parietal cortex support episodic memory? Evidence from focal lesion patients.Neuropsychologia461743–1755. 10.1016/j.neuropsychologia.2008.01.011
34
Della-MaggioreV.SekulerA. B.GradyC. L.BennettP. J.SekulerR.McIntoshA. R. (2000). Corticolimbic interactions associated with performance on a short-term memory task are modified by age.J. Neurosci.208410–8416. 10.1523/jneurosci.20-22-08410.2000
35
di PellegrinoG.CiaramelliE.LàdavasE. (2007). The regulation of cognitive control following rostral anterior cingulate cortex lesion in humans.J. Cogn. Neurosci.19275–286. 10.1162/jocn.2007.19.2.275
36
DobbinsI. G.FoleyH.SchacterD. L.WagnerA. D. (2002). Executive control during episodic retrieval: multiple prefrontal processes subserve source memory.Neuron35989–996. 10.1016/s0896-6273(02)00858-9
37
DudukovicN. M.WagnerA. D. (2007). Goal-dependent modulation of declarative memory: neural correlates of temporal recency decisions and novelty detection.Neuropsychologia452608–2620. 10.1016/j.neuropsychologia.2007.02.025
38
DzaficI.OestreichL.MartinA. K.MowryB.BurianovàH. (2019). Stria terminalis, amygdala, and temporoparietal junction networks facilitate efficient emotion processing under expectations.Hum. Brain Mapp.405382–5396. 10.1002/hbm.24779
39
EdgingtonE. S. (1980). Randomization Tests.New York, NY: Marcel Dekker.
40
EfronB.TibshiraniR. (1985). The bootstrap method for assessing statistical accuracy.Behaviormetrika171–35. 10.2333/bhmk.12.17_1
41
FavilaS. E.SamideR.SweigartS. C.KuhlB. A. (2018). Parietal representations of stimulus features are amplified during memory retrieval and flexibly aligned with top-down goals.J. Neurosci.387809–7821. 10.1523/jneurosci.0564-18.2018
42
FleckM. S.DaselaarS. M.DobbinsI. G.CabezaR. (2006). Role of prefrontal and anterior cingulate regions in decision-making processes shared by memory and non-memory tasks.Cereb. Cortex161623–1630. 10.1093/cercor/bhj097
43
FoxK. C.SprengR. N.EllamilM.Andrews-HannaJ. R.ChristoffK. (2015). The wandering brain: meta-analysis of functional neuroimaging studies of mind-wandering and related spontaneous thought processes.Neuroimage111611–621. 10.1016/j.neuroimage.2015.02.039
44
FristonK. J.FrithC. D.LiddleP. F.FrackowiakR. S. J. (1993). Functional connectivity: the principal component analysis of large (PET) data sets.J. Cereb. Blood Flow Metab.135–14. 10.1038/jcbfm.1993.4
45
GilmoreA. W.NelsonS. M.McDermottK. B. (2015). A parietal memory network revealed by multiple MRI methods.Trends Cogn. Sci.19534–543. 10.1016/j.tics.2015.07.004
46
GuerinS. A.MillerM. B. (2011). Parietal cortex tracks the amount of information retrieved even when it is not the basis of a memory decision.Neuroimage55801–807. 10.1016/j.neuroimage.2010.11.066
47
HallS. A.RubinD. C.MilesA.DavisS. W.WingE. A.CabezaR.et al (2014). The neural basis of involuntary episodic memories.J. Cogn. Neurosci.262385–2399. 10.1162/jocn_a_00633
48
HanS.HuettelS. A.RaposoA.AdcockR. A.DobbinsI. G. (2010). Functional significance of striatal responses during episodic decisions: recovery or goal attainment?J. Neurosci.304767–4775. 10.1523/jneurosci.3077-09.2010
49
HayesS. M.BuchlerN.StokesJ.KragelJ.CabezaR. (2011). Neural correlates of confidence during item recognition and source memory retrieval: evidence for both dual-process and strength memory theories.J. Cogn. Neurosci.233959–3971. 10.1162/jocn_a_00086
50
HealyA. F.KubovyM. (1978). The effects of payoffs and prior probabilities on indices of performance and cut off location in recognition memory.Mem. Cogn.6544–553. 10.3758/bf03198243
51
HembacherE.GhettiS. (2014). Don’t look at my answer: subjective uncertainty underlies preschoolers’ exclusion of their least accurate memories.Psychol. Sci.251768–1776. 10.1177/0956797614542273
52
HensonR. N.ShalliceT.DolanR. J. (1999). Right prefrontal cortex and episodic memory retrieval: a functional MRI test of the monitoring hypothesis.Brain1221367–1381. 10.1093/brain/122.7.1367
53
HerronJ. E.HensonR. N.RuggM. D. (2004). Probability effects on the neural correlates of retrieval success: an fMRI study.Neuroimage21302–310. 10.1016/j.neuroimage.2003.09.039
54
HumphreysG. F.Lambon RalphM. A. (2017). Mapping domain-selective and counterpointed domain-general higher cognitive functions in the lateral parietal cortex: evidence from fMRI comparisons of difficulty-varying semantic versus visuo-spatial tasks, and functional connectivity analyses.Cereb. Cortex274199–4212. 10.1093/cercor/bhx107
55
HutchinsonJ. B.UncapherM. R.WagnerA. D. (2009). Posterior parietal cortex and episodic retrieval: convergent and divergent effects of attention and memory.Learn. Mem.16343–356. 10.1101/lm.919109
56
HutchinsonJ. B.UncapherM. R.WeinerK. S.BresslerD. W.SilverM. A.PrestonA. R.et al (2014). Functional heterogeneity in posterior parietal cortex across attention and episodic memory retrieval.Cereb. Cortex2449–66. 10.1093/cercor/bhs278
57
JacobyL. L.JonesT. C.DolanP. O. (1998). Two effects of repetition: support for a dual-process model of know judgments and exclusion errors.Psychol. Bull. Rev.5705–709. 10.3758/bf03208849
58
JaegerA.KonkelA.DobbinsI. G. (2013). Unexpected novelty and familiarity orienting responses in lateral parietal cortex during recognition judgment.Neuropsychologia511061–1076. 10.1016/j.neuropsychologia.2013.02.018
59
KimH.CabezaR. (2007). Trusting our memories: dissociating the neural correlates of confidence in veridical versus illusory memories.J. Neurosci.2712190–12197. 10.1523/jneurosci.3408-07.2007
60
KingD. R.MillerM. B. (2017). Influence of response bias and internal/external source on lateral posterior parietal successful retrieval activity.Cortex91126–141. 10.1016/j.cortex.2017.04.002
61
KrishnanA.WilliamsL. J.McIntoshA. R.AbdiH. (2011). Partial Least Squares (PLS) methods for neuroimaging: a tutorial and review.Neuroimage56455–475. 10.1016/j.neuroimage.2010.07.034
62
KuceraH.FrancisW. N. (1967). Computational Analysis of Present-Day American English.Providence, RI: Brown University Press.
63
KuhlB. A.JohnsonM. K.ChunM. M. (2013). Dissociable neural mechanisms for goal-directed versus incidental memory reactivation.J. Neurosci.3316099–16109. 10.1523/jneurosci.0207-13.2013
64
KumaranD.MaguireE. A. (2006). An unexpected sequence of events: mismatch detection in the human hippocampus.PLoS Biol.4:e424. 10.1371/journal.pbio.0040424
65
LakensD. (2017). Equivalence tests: a practical primer for t-tests, correlations, and meta-analyses.Soc. Psychol. Pers. Sci.8355–362. 10.1177/1948550617697177
66
LaMontagneP. J.HabibR. (2010). Stimulus-driven incidental episodic retrieval involves activation of the left posterior parietal cortex.Neuropsychologia483317–3322. 10.1016/j.neuropsychologia.2010.07.015
67
MacmillanN. A.CreelmanC. D. (1991). Detection Theory: A User’s Guide.New York, NY: Cambridge University Press.
68
MaroisR.LeungH. C.GoreJ. C. (2000). A stimulus-driven approach to object identity and location processing in the human brain.Neuron25717–728. 10.1016/s0896-6273(00)81073-9
69
MarstallerL.BurianováH. (2015). A common functional neural network for overt production of speech and gesture.Neuroscience28429–41. 10.1016/j.neuroscience.2014.09.067
70
MartinV. C.SchacterD. L.CorballisM. C.AddisD. R. (2011). A role for the hippocampus in encoding simulations of future events.Proc. Natl. Acad. Sci. U.S.A.3313858–13863. 10.1073/pnas.1105816108
71
McCormickC.St-LaurentM.TyA.ValianteT. A.McAndrewsM. P. (2015). Functional and effective hippocampal-neocortical connectivity during construction and elaboration of autobiographical memory retrieval.Cereb. Cortex251297–1305. 10.1093/cercor/bht324
72
McIntoshA. R. (1999). Mapping cognition to the brain through neural interactions.Memory7523–548. 10.1080/096582199387733
73
McIntoshA. R.Gonzalez-LimaF. (1994). Structural equation modeling and its application to network analysis in functional brain imaging.Hum. Brain Mapp.22–22. 10.1002/hbm.460020104
74
McIntoshA. R.LobaughN. J. (2004). Partial least squares analysis of neuroimaging data: applications and advances.Neuroimage23 (Suppl. 1), S250–S263.
75
McIntoshA. R.NybergL.BooksteinF. L.TulvingE. (1997). Differential functional connectivity of prefrontal and medial temporal cortices during episodic memory retrieval.Hum. Brain Mapp.5323–327. 10.1002/(sici)1097-0193(1997)5:4<323::aid-hbm20>3.0.co;2-d
76
MillerM.HandyT. C.CutlerJ.InatiS.WolfordG. L. (2001). Brain activations associated with shifts in response criterion on a recognition test.Can. J. Exp. Psychol.55162–173. 10.1037/h0087363
77
MoscovitchM. (2008). The hippocampus as a “stupid,” domain-specific module: implications for theories of recent and remote memory, and of imagination.Can. J. Exp. Psychol.6262–79. 10.1037/1196-1961.62.1.62
78
MoscovitchM.CabezaR.WinocurG.NadelL. (2016). Episodic memory and beyond: the hippocampus and neocortex in transformation.Ann. Rev. Psychol.67105–134. 10.1146/annurev-psych-113011-143733
79
MoscovitchM.WinocurG. (1992). “The neuropsychology of memory and aging,” in The Handbook of Aging and Cognition, edsCraikF. I. M.SalthouseT. A. (Hillsdale, NJ: Lawrence Erlbaum Associates).
80
NelsonS. M.CohenA. L.PowerJ. D.WigG. S.MiezinF. M.WheelerM. E.et al (2010). A parcellation scheme for human left lateral parietal cortex.Neuron67156–170. 10.1016/j.neuron.2010.05.025
81
O’ConnorA. R.HanS.DobbinsI. G. (2010). The inferior parietal lobule and recognition memory: expectancy violation or successful retrieval?J. Neurosci.302924–2934. 10.1523/jneurosci.4225-09.2010
82
OgawaS.MenonR. S.TankD. W.KimS. G.MerkleH.EllermannJ. M.et al (1993). Functional brain mapping by blood oxygenation level-dependent contrast magnetic resonance imaging. A comparison of signal characteristics with a biophysical model.Biophys. J.64803–812. 10.1016/s0006-3495(93)81441-3
83
PlattM. L.GlimcherP. W. (1999). Neural correlates of decision variables in parietal cortex.Nature400233–238. 10.1038/22268
84
PoppenkJ.EvensmoenH. R.MoscovitchM.NadelL. (2013). Long-axis specialization of the human hippocampus.Trends Cogn. Sci.17230–240. 10.1016/j.tics.2013.03.005
85
PoppenkJ.MoscovitchM. (2011). A hippocampal marker of recollection memory ability among healthy young adults: contributions of posterior and anterior segments.Neuron72931–937. 10.1016/j.neuron.2011.10.014
86
PrestonA. R.EichenbaumH. (2013). Interplay of hippocampus and prefrontal cortex in memory.Curr. Biol.23R764–R773.
87
QuammeJ. R.WeissD. J.NormanK. A. (2010). Listening for recollection: a multi-voxel pattern analysis of recognition memory retrieval strategies.Front. Hum. Neurosci.4:61. 10.3389/fnhum.2010.00061
88
RamananS.PiguetO.IrishM. (2018). Rethinking the role of the angular gyrus in remembering the past and imagining the future: the contextual integration model.Neuroscientist24342–352. 10.1177/1073858417735514
89
RanganathC.JohnsonM. K.D’EspositoM. (2000). Left anterior prefrontal activation increases with demands to recall specific perceptual information.J. Neurosci.20:RC108.
90
RobinJ.HirshhornM.RosenbaumR. S.WinocurG.MoscovitchM.GradyC. L. (2015). Functional connectivity of hippocampal and prefrontal networks during episodic and spatial memory based on real-world environments.Hippocampus2581–93. 10.1002/hipo.22352
91
RobinJ.MoscovitchM. (2017). Details, gist and schema: hippocampal-neocortical interactions underlying recent and remote episodic and spatial memory.Curr. Opin. Behav. Sci.17114–123. 10.1016/j.cobeha.2017.07.016
92
RossiS.CappaS. F.BabiloniC.PasqualettiP.MiniussiC.CarducciF.et al (2001). Prefrontal cortex in long-term memory: an “interference” approach using magnetic stimulation.Nat. Neurosci.4948–952. 10.1038/nn0901-948
93
RuggM. D.FletcherP. C.ChuaP. M.DolanR. J. (1999). The role of the prefrontal cortex in recognition memory and memory for source: an fMRI study.Neuroimage10520–529. 10.1006/nimg.1999.0488
94
RuggM. D.KingD. R. (2018). Ventral lateral parietal cortex and episodic memory retrieval.Cortex107238–250. 10.1016/j.cortex.2017.07.012
95
RuggM. D.VilbergK. L. (2013). Brain networks underlying episodic memory retrieval.Curr. Opin. Neurobiol.23255–260. 10.1016/j.conb.2012.11.005
96
RushworthM. F. S.EllisonA.WalshA. (2001). Complementary localization and lateralization of orienting and motor attention.Nat. Neurosci.4656–661. 10.1038/88492
97
SampsonP. D.StreissguthA. P.BarrH. M.BooksteinF. L. (1989). Neuro-behavioral effects of prenatal alcohol: Part II. Partial least-squares analysis.Neurotoxicol. Teratol.11477–491. 10.1016/0892-0362(89)90025-1
98
SchreursB.McIntoshA. R.BahronM.HerscovitchP.SunderlandT.MolchanS. (1997). Lateralization and behavioural correlation of changes in regional cerebral blood flow with classical conditioning of the human eyeblink response.J. Neurophysiol.772153–2163. 10.1152/jn.1997.77.4.2153
99
ScimecaJ. M.BadreD. (2012). Striatal contributions to declarative memory retrieval.Neuron73380–392. 10.1016/j.neuron.2012.07.014
100
SeeleyW. W.MenonV.SchatzbergA. F.KellerJ.GloverG. H.KennaH.et al (2007). Dissociable intrinsic connectivity networks for salience processing and executive control.J. Neurosci.272349–2356. 10.1523/jneurosci.5587-06.2007
101
SestieriC.ShulmanG. L.CorbettaM. (2010). Attention to memory and the environment: functional specialization and dynamic competition in human posterior parietal cortex.J. Neurosci.238445–8456. 10.1523/jneurosci.4719-09.2010
102
SestieriC.ShulmanG. L.CorbettaM. (2017). The contribution of the human posterior parietal cortex to episodic memory.Nat. Rev. Neurosci.18183–192. 10.1038/nrn.2017.6
103
ShimamuraA. P. (2011). Episodic retrieval and the cortical binding of relational activity.Cogn. Affect. Behav. Neurosci.11277–291. 10.3758/s13415-011-0031-4
104
SimonsJ. S.PeersP. V.MazuzY. S.BerryhillM. E.OlsonI. R. (2010). Dissociation between memory accuracy and memory confidence following bilateral parietal lesions.Cereb. Cortex20479–485. 10.1093/cercor/bhp116
105
SimonsJ. S.SpiersH. J. (2003). Prefrontal and medial temporal lobe interactions in long-term memory.Nat. Rev. Neurosci.4637–648. 10.1038/nrn1178
106
StellaF.CerastiE.SiB.JezekK.TrevesA. (2012). Self-organization of multiple spatial and context memories in the hippocampus.Neurosci. Biobehav. Rev.361609–1625. 10.1016/j.neubiorev.2011.12.002
107
St-LaurentM.HervèA.BurianovaH.GradyC. (2011). Influence of aging on the neural correlates of autobiographical, episodic, and semantic memory retrieval.J. Cogn. Neurosci.234150–4163. 10.1162/jocn_a_00079
108
StrangeB. A.WitterM. P.LeinE. S.MoserE. I. (2014). Functional organization of the hippocampal longitudinal axis.Nat. Rev. Neurosci.15655–669. 10.1038/nrn3785
109
StretchV.WixtedJ. T. (1998). On the difference between strength-based and frequency-based mirror effects in recognition memory.J. Exp. Psychol: Learn. Mem. Cogn.241379–1396. 10.1037/0278-7393.24.6.1379
110
TeylerT. J.RudyJ. W. (2007). The hippocampal indexing theory and episodic memory: updating the index.Hippocampus171158–1169. 10.1002/hipo.20350
111
UddinL. Q.SupekarK.AminH.RykhlevskaiaE.NguyenD. A.GreiciusM. D.et al (2010). Dissociable connectivity within human angular gyrus and intraparietal sulcus: evidence from functional and structural connectivity.Cereb. Cortex202636–2646. 10.1093/cercor/bhq011
112
VallesiA.McIntoshA. R.AlexanderM.StussD. T. (2009). FMRI evidence of a functional network setting the criteria for withholding a response.Neuroimage45537–548. 10.1016/j.neuroimage.2008.12.032
113
VilbergK. L.RuggM. D. (2007). Dissociation of the neural correlates of recognition memory according to familiarity, recollection, and amount of recollected information.Neuropsychologia452216–2225. 10.1016/j.neuropsychologia.2007.02.027
114
VilbergK. L.RuggM. D. (2009). An investigation of the effects of relative probability of old and new test items on the neural correlates of successful and unsuccessful source memory.Neuroimage45562–571. 10.1016/j.neuroimage.2008.12.020
115
WagnerA. D.ShannonB. J.KahnI.BucknerR. L. (2005). Parietal lobe contributions to episodic memory retrieval.Trends Cogn. Sci.9445–453. 10.1016/j.tics.2005.07.001
116
YazarY.BergstromZ. M.SimonsJ. S. (2017). Reduced multimodal integration of memory features following continuous theta burst stimulation of angular gyrus.Brain Stimul.10624–629. 10.1016/j.brs.2017.02.011
117
ZiaeiM.SalamiA.PerssonJ. (2017). Age-related alterations in functional connectivity patterns during working memory encoding of emotional items.Neuropsychologia941–12. 10.1016/j.neuropsychologia.2016.11.012
118
ZweynertS.PadeJ. P.WüstenbergT.SterzerP.WalterH.SeidenbecherC. I.et al (2011). Motivational salience modulates hippocampal repetition suppression and functional connectivity in humans.Front. Hum. Neurosci.5:144. 10.3389/fnhum.2011.00144
Summary
Keywords
episodic memory, recognition memory decision, posterior parietal cortex, hippocampus, functional magnetic brain imaging (fMRI)
Citation
Ciaramelli E, Burianová H, Vallesi A, Cabeza R and Moscovitch M (2020) Functional Interplay Between Posterior Parietal Cortex and Hippocampus During Detection of Memory Targets and Non-targets. Front. Neurosci. 14:563768. doi: 10.3389/fnins.2020.563768
Received
19 May 2020
Accepted
16 October 2020
Published
03 November 2020
Volume
14 - 2020
Edited by
Monica Luciana, University of Minnesota Twin Cities, United States
Reviewed by
Christine Bastin, University of Liège, Belgium; G. Elliott Wimmer, University College London, United Kingdom
Updates
Copyright
© 2020 Ciaramelli, Burianová, Vallesi, Cabeza and Moscovitch.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Elisa Ciaramelli, elisa.ciaramelli@unibo.it
This article was submitted to Decision Neuroscience, a section of the journal Frontiers in Neuroscience
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.