Are There Multiple Kinds of Episodic Memory PDF

2764 The Journal of Neuroscience, March 8, 2017 37(10):2764 –2775 Behavioral/Cognitive Are There Multiple Kinds of Episodic Memory? An fMRI Investigation Comparing Autobiographical and Recognition Memory Tasks X Hung-Yu Chen,1 Adrian W. Gilmore,1 Steven M. Nelson,2,3,4 and X Kathleen B. McDerm...

2764 The Journal of Neuroscience, March 8, 2017 37(10):2764 –2775 Behavioral/Cognitive Are There Multiple Kinds of Episodic Memory? An fMRI Investigation Comparing Autobiographical and Recognition Memory Tasks X Hung-Yu Chen,1 Adrian W. Gilmore,1 Steven M. Nelson,2,3,4 and X Kathleen B. McDermott1,5 1Department of Psychological and Brain Sciences, Washington University in St. Louis, St. Louis, Missouri, 63130, 2VISN 17 Center of Excellence for Research on Returning War Veterans, Waco, Texas, 76711, 3Center for Vital Longevity, School of Behavioral and Brain Sciences, University of Texas at Dallas, Dallas, Texas 75235, 4Department of Psychology and Neuroscience, Baylor University, Waco, Texas 76798, and 5Department of Radiology, Washington University in St. Louis, St. Louis, Missouri, 63110 What brain regions underlie retrieval from episodic memory? The bulk of research addressing this question with fMRI has relied upon recognition memory for materials encoded within the laboratory. Another, less dominant tradition has used autobiographical methods, whereby people recall events from their lifetime, often after being cued with words or pictures. The current study addresses how the neural substrates of successful memory retrieval differed as a function of the targeted memory when the experimental parameters were held constant in the two conditions (except for instructions). Human participants studied a set of scenes and then took two types of memory test while undergoing fMRI scanning. In one condition (the picture memory test), participants reported for each scene (32 studied, 64 nonstudied) whether it was recollected from the prior study episode. In a second condition (the life memory test), participants reported for each scene (32 studied, 64 nonstudied) whether it reminded them of a specific event from their preexperimental lifetime. An examination of successful retrieval (yes responses) for recently studied scenes for the two test types revealed pronounced differences; that is, autobiographical retrieval instantiated with the life memory test preferentially activated the default mode network, whereas hits in the picture memory test preferentially engaged the parietal memory network as well as portions of the frontoparietal control network. When experimental cueing parameters are held constant, the neural underpinnings of successful memory retrieval differ when remembering life events and recently learned events. Key words: autobiographical memory; default mode network; episodic memory; fMRI; parietal memory network; recognition memory Significance Statement Episodic memory is often discussed as a solitary construct. However, experimental traditions examining episodic memory use very different approaches, and these are rarely compared to one another. When the neural correlates associated with each ap- proach have been directly contrasted, results have varied considerably and at times contradicted each other. The present experi- ment was designed to match the two primary approaches to studying episodic memory in an unparalleled manner. Results suggest a clear separation of systems supporting memory as it is typically tested in the laboratory and memory as assessed under autobiographical retrieval conditions. These data provide neurobiological evidence that episodic memory is not a single construct, challenging the degree to which different experimental traditions are studying the same construct. Introduction broad research paths. In the more common approach, which has The investigation of episodic memory, defined as memory for events been termed the laboratory tradition (McDermott et al., 2009; or episodes in one’s past (Tulving, 1972), has proceeded along two Roediger and McDermott, 2013), participants learn and retrieve vided by Jonathan Power, from assistance with data collection by Fan Zou (Experiment 1) and Ruthie Shaffer and Received May 10, 2016; revised Jan. 19, 2017; accepted Jan. 26, 2017. Hannah Becker (Experiment 2), and from response scoring by Jiayi Zhou (Experiment 1). Author contributions: H.-Y.C., A.W.G., S.M.N., and K.B.M. designed research; H.-Y.C. and A.W.G. performed The authors declare no competing financial interests. research; H.-Y.C., A.W.G., and K.B.M. analyzed data; H.-Y.C., A.W.G., and K.B.M. wrote the paper. Correspondence should be addressed to Kathleen McDermott at the above address. E-mail: This work was supported by grants from the McDonnell Center for Systems Neuroscience at Washington Univer- [email protected]. sity and Dart NeuroScience, LLC, and by the NSF Graduate Research Fellowship Program (DGE-1143954 to A.W.G.). DOI:10.1523/JNEUROSCI.1534-16.2017 This work benefitted from discussions with Ian Dobbins and Jeff Zacks, from network parcellation underlays pro- Copyright © 2017 the authors 0270-6474/17/372764-12$15.00/0 Chen et al. Two Types of Episodic Memory Identified with fMRI J. Neurosci., March 8, 2017 37(10):2764 –2775 2765 materials within an experimental context. This approach affords autobiographical conditions demonstrate that even within this considerable experimenter control, and it assumes that similar rules task setting, and within a single group of subjects, there are clear will govern the remembrance of any single experience, whether a differences in the neural underpinnings of the two types of word on a screen or a meal with a friend (Tulving, 1972). An alter- memory. native approach rejects this assumption, and is guided by the belief that to properly understand memory for lifetime events, one must forgo the control offered by list-learning tasks and instead ask par- Materials and Methods ticipants to retrieve memories from their daily lives, as would be Experiment 1: fMRI study Participants. Thirty-one participants (16 female, ages 18 –35) were re- done in more “natural” settings. cruited from Washington University and the St. Louis area. Participants Are the two methodologies tapping fundamentally similar were all right-handed, native speakers of English (acquired by the age of memory systems, or is it necessary to use autobiographical meth- 5), had normal or corrected-to-normal vision, and were free of psychi- ods if the goal is to understand retrieval of events from one’s atric or neurological disorders. Data from three participants were lifetime? Posing this question is difficult with purely behavioral excluded from analysis due to experimenter error. In addition, one par- measures. A more direct way of approaching the question is to ticipant was excluded due to excessive motion, leaving a final sample of use functional neuroimaging methods, in which putative func- N ⫽ 27. Informed consent was obtained for all participants, and the tions can be ascribed to specific regions of the brain. Relatively study was conducted in accordance with Washington University human few studies have attempted to address the question of whether research practices. Participants were paid $25 per hour. Materials. Scenes were chosen as cues for two reasons. First, pilot work (and how) the different traditions of psychological memory re- demonstrated that scenes were effective cues for autobiographical mem- search may be similarly or differentially supported in the brain, ory (as effective or more effective than words). Second, the average re- and they have reached disparate (and at times conflicting) con- sponse time (RT) for scenes to elicit autobiographical memories fell clusions. Some evidence suggests that regions within the medial within 1 s of the response time for recognition memory (a much smaller temporal lobe and medial prefrontal cortex exhibit greater activ- difference than when words are used as cues). ity during autobiographical memory retrieval than during recog- Collection of scene stimuli followed the procedure used by Konkle et nition memory (Conway et al., 1999; Nyberg et al., 2002; Cabeza al. (2010): 284 images of various categories (such as a cafeteria, a lecture et al., 2004; Hassabis et al., 2007; Summerfield et al., 2009; Elman hall, tennis courts, and an airport) were gathered via Google Images et al., 2013). Evidence of regions showing the opposite pattern is (images.google.com). Only images with resolution higher than 800 ⫻ less consistent, with only three of the previously cited studies 600 pixels were used. None of the scenes contained people. The scenes were then resized to 400 ⫻ 300 pixels (overall screen resolution: 1024 ⫻ locating such effects across frontal and parietal regions (Conway 768 pixels). Scenes were rotated across conditions across participants. et al., 1999; Nyberg et al., 2002; Elman et al., 2013). It may there- Encoding and retrieval instructions and procedures. The procedure of fore be the case that similar neural substrates support episodic the experiment is portrayed in Figure 1. All of the tasks took place in the memory broadly, with autobiographical memory retrieval driv- scanner, and participants remained in the scanner between encoding and ing such regions to greater activity levels (possibly due to greater retrieval. Participants began by studying 126 indoor or outdoor scenes; complexity). This hypothesis invokes (often implicitly) the as- each scene was displayed for 2 s, followed by a blank screen for 500 ms. sumptions of the verbal learning tradition, in which word lists Participants used a button press to classify each scene as indoor or out- and similar materials are thought to serve as a proxy for life events door while learning the scenes (intentional encoding). (Tulving, 1983). During the memory test requiring retrieval from autobiographical A recent meta-analysis suggests a different conclusion, how- memory, participants were told that they were taking a life memory test and were asked to report (via button press) whether they could use the ever. Specifically, an activation likelihood estimation procedure picture “to help you remember a specific event or moment in your life” was used to quantify regions of the brain that tend to be active and to “please try hard to remember specific details about an event from during autobiographical memory tasks (relative to a variety of your life that is distinct in time and place.” control conditions) and those that tend to be active during suc- In the memory test requiring retrieval of recently learned stimuli, cessful recognition memory (comparing activity for hits to cor- participants were told that they were taking a picture memory test and rectly rejected lures); the resulting voxelwise maps were almost were asked to report (via button press) whether they could “remember nonoverlapping (McDermott et al., 2009). This analytic ap- having seen” the picture in the study phase. Furthermore, they were proach suggests that recognition memory and autobiographical asked for each picture to “try hard to remember specific details about memory recruit distinct regions of cortex and therefore measure having seen it before (e.g., some specific feature of the picture that you distinct types of memory. recall looking at, or what the picture made you think of when you studied it.” Hence, both instructions emphasized the importance of remember- Is this conclusion justified, or might differences in the proce- ing, recollecting, or reliving to the extent possible (Tulving, 1985), and dures (e.g., different stimulus cues, different retention intervals, the instructions were kept as similar as possible (given that the instruc- different response times) inherent in the two literatures lead to tions constituted the experimental manipulation). Participants went this outcome? The purpose of the present experiment was to through a few training trials, and they did not enter the scanner until it directly compare the neural correlates supporting memory re- was clear to the experimenter that they understood the instructions. Half trieval within these traditions, striking a balance between exper- of the participants began with the life memory test, whereas the other half imental control and ecological validity. Participants encoded began with the picture memory test. Those subjects receiving the picture scenes and later viewed a mixture of new and old scenes, and were memory test first began the test approximately 1 min after the end of the asked about their memories in response to each stimulus. To encoding run. The remaining participants began the test ⬃16 min after represent the laboratory-based tradition, participants made yes/no the end of encoding. At the end of the scanning session, all participants completed an additional autobiographical retrieval task (longer trials, no responses indicating whether a scene had been recently encoded. button press, modeled after Szpunar et al., 2007); data from that task are To represent the autobiographical tradition, participants made not reported here. yes/no responses to indicate whether or not a stimulus reminded Both types of tests were divided into two blocks (or runs) of 48 trials them of an event from earlier in their lifetime. All variables, in- (16 old or “studied” scenes, 32 new or “nonstudied” scenes). For each cluding type of stimuli presented, test item history, and trial du- trial, a scene appeared on the screen for 4 s and was followed by a blank screen ration, were matched closely. Direct contrasts of laboratory and lasting 1 s. Participants were given 5 s from trial onset to respond with a 2766 J. Neurosci., March 8, 2017 37(10):2764 –2775 Chen et al. Two Types of Episodic Memory Identified with fMRI Figure 1. Experimental procedures. Participants viewed 126 scenes (intentional encoding) with a binary (indoor/outdoor) judgment for each. They then took two types of memory tests: recognition and autobiographical (counterbalanced order). Each test consisted of 96 scenes (48 scenes in each of two runs; totaling 64 new, 32 old) for 4 s with a 1 s blank screen following. The only difference between test type was the instructions and the title given to subjects (picture memory test or life memory test). Images in this figure are publicly available on Wikimedia Commons (attributions from left to right: Kevin Payravi, Kapacytron, Foxparabola). ISI, Interstimulus interval. button press to indicate whether they recognized the scene or whether the series of affine transforms (Michelon et al., 2003). Individual subject data scene could be used to retrieve a specific life event (for the picture memory were averaged across people to create a group average contrast map. The test and life memory test, respectively). In sum, item history, stimulus tim- contrast images were smoothed using a Gaussian smoothing kernel with ing, and response output modality were identical for the two tests. 6 mm FWHM. fMRI data acquisition. Functional and structural scans were acquired GLM coding. Each memory test run consisted of 157 frames (TRs), on a Siemens 3.0T MAGNETOM Trio system using a Siemens 12- although the first four were dropped, leaving 153 frames analyzed per channel head coil. Stimuli were presented with PsyScope (Cohen, 1993) run. One run from one participant was dropped due to within-run on an iMac computer, which received sync pulses from the scanner. movement. Participants’ individual runs were concatenated into a single Length of jitter and randomization of trial types were optimized using the time series. program Optseq2 (http://surfer.nmr.mgh.harvard.edu/optseq/). The data were modeled with a general linear model, which included Structural images were acquired using a T1-weighted sagittal MPRAGE eight regressors of interest: four corresponding to recognition memory (TE, 3.08 ms; TR (partition), 2.4 s; TI, 1000 ms; flip angle, 8°; 176 slices trial types (hits, misses, correct rejections, and false alarms) and four with resolution 1 ⫻ 1 ⫻ 1 mm voxels), and were used along with a corresponding to autobiographical memory trial types (successful re- T2-weighted turbo spin echo structural image (TE, 84 ms; TR, 6.8 s; 32 trieval for old scenes, unsuccessful retrieval for old scenes, successful slices with 2 ⫻ 1 ⫻ 4 mm voxels) in the same anatomical plane as the retrieval for new scenes, and unsuccessful retrieval for new scenes). For BOLD images to improve atlas alignment. each participant, RTs for each trial were z scored and included as a Gradient field maps allowed estimation of inhomogeneities in the regressor. We analyzed the data with and without these RT regressors, magnetic field for each subject. An autoalign pulse sequence protocol and the contrast between successful autobiographical retrieval and suc- provided in the Siemens software was used to align the acquisition slices cessful picture memory did not vary appreciably; analyses reported here of the functional scans parallel to the anterior commissure–posterior included RT as a regressor. Regressors of no interest included a trend commissure plane. Slices collected were therefore parallel to the slices in term to account for linear changes and a constant term to model the the Talairach atlas (Talairach and Tournoux, 1988). Functional imaging baseline. A standard hemodynamic response function was chosen (Boy- used a BOLD contrast sensitive gradient-echo echoplanar sequence (TE, nton et al., 1996) to estimate the hemodynamic response for each condi- 27 ms; flip angle, 90°; in-plane resolution, 4 ⫻ 4 mm). Whole-brain EPI tion, with an onset delay of 2 s. Effects were analyzed in terms of percent volumes (MR frames) of 32 contiguous, 4-mm-thick axial slices were signal change relative to baseline. obtained every 2.5 s. The first four functional images of each scan were Analysis and visualization software. Imaging analysis was done using discarded to allow for T1 equilibration effects. Washington University’s in-house software, FIDL (http://nil.wustl.edu/ fMRI data preprocessing. Imaging data from each participant were pre- ~fidl). All reported atlas coordinates were converted from 711-2C space processed to remove noise and artifacts including (1) temporal realign- to MNI 152 space. Figures displaying statistical maps were made by pro- ment using sinc interpolation of all slices to the temporal midpoint of the jecting and displaying the volumetric data onto a partially inflated rep- first slice to account for differences in slice time acquisition, (2) correc- resentation of the human brain using the Connectome Workbench tion for movement within and across scan runs using a rigid-body rota- software (Marcus et al., 2011). tion and translation algorithm (Snyder, 1996), (3) gradient field map Retrieval tasks voxelwise t test analysis and ROI definition. The obtained correction to correct for spatial distortion due to local field inhomoge- t test images were multiple-comparison-corrected to a whole-brain fami- neities using FMRIB Software Library’s FUGUE (http://fsl.fMRIb.ox.ac. lywise error rate of p ⬍ 0.05 using a z ⬎ 3 threshold with at least 17 uk), and (4) whole-brain normalization with a single constant factor to a contiguous voxels (McAvoy et al., 2001). The cluster size threshold T common mode of 1000 in the fifth frame to allow for comparisons across values were chosen based on 10,000 Monte Carlo simulations performed participants (Ojemann et al., 1997). Functional data were then resampled by McAvoy et al. (2001). This cluster-level correction was designed to using 3 mm isotropic voxels and transformed into stereotaxic atlas space provide adequate control for false positives without a significant loss of (Talairach and Tournoux, 1988). Atlas registration involved aligning statistical power. An automated algorithm (peak_4dfp) written by A. each participant’s T1-weighted image to a custom atlas-transformed Snyder (Washington University School of Medicine) searched for the loca- (Lancaster et al., 1995) target T1-weighted template (711-2C) using a tion of peaks in the resulting image and drew spheres (16 mm diameter) Chen et al. Two Types of Episodic Memory Identified with fMRI J. Neurosci., March 8, 2017 37(10):2764 –2775 2767 Figure 2. Accuracy and response latency for the two types of tests. A, Mean proportions of “yes” and “no” responses to old (studied) and new (nonstudied) scenes for the life memory test (red) and the picture memory test (blue). B, Mean response latencies for “yes” and “no” responses to old (studied) and new (nonstudied) scenes for the life memory test (in red) and the picture memory test (blue). Error bars display SEM. Figure 3. Differential activation for successful recollection in the life memory test and picture memory test. A, A contrast of old scenes that elicited “yes” responses to indicate successful retrieval. B, A contrast of new scenes that elicited “yes” responses in the life memory test and old scenes that elicited “yes” responses in the picture memory test. For both panels, z ⬎ 3, k ⬎ 17, whole-brain p ⬍ 0.05 (familywise error corrected). around each peak. Peaks under 16 mm apart were consolidated via coordi- sponse equated), the latter contrast is more like what one would achieve nate averaging. Regions of interest (ROIs) were then obtained by masking in contrasting the conditions typically used to study episodic retrieval. the 16 mm spheres by the multiple comparison corrected image. Regions We primarily focus on the prior contrast due to the better match in item located in white matter, CSF, or ventricles were excluded from analysis. history. In addition, a contrast between successful retrieval in autobio- The primary advantage of this experimental design was that we could graphical memory for studied and nonstudied scenes was analyzed to contrast conditions with identical item history (e.g., old scenes) and assess possible “contamination” of recognition signal in autobiographi- identical overt response (“yes, I remember”) but differing underlying cal retrieval. Also shown (for completeness) are one-tailed t tests of both experiences (remembering having studied the scene or using the scene to test types (all items) relative to the low-level control of the implicit base- remember a preexperimental life event); that is, the primary contrast of line. Finally, we report contrasts between hits and correctly rejected lures interest was a voxelwise t test (paired sample, two-tailed) contrasting in the picture memory test and between old scenes triggering successful activity estimates for successful retrieval cued by previously studied memory retrieval and new scenes not triggering memory retrieval in the scenes within the life memory test and the picture memory test. The life memory test. contrast compared hits on a recognition memory test to successful re- Retrieval tasks network-wide comparison. Results from the whole-brain trieval on an autobiographical memory test (where recently studied analyses described above suggested a network-wide dissociation for suc- scenes were the cues). cessful retrieval in the picture memory test and the life memory test. To Other contrasts of interest included a comparison similar to the one further explore this possibility, we used the 264 ROIs reported by a prior above, but where novel, nonstudied scenes served as the items contrasted whole-brain network parcellation study (Power et al., 2011) to examine to recognition hits. This contrast is more analogous to a direct compar- individual networks’ activity during successful memory for previously ison between the conditions typically used in the literature; that is, studied scenes (“yes” responses in the picture memory test and life whereas the prior contrast is more elegant (with item history and re- memory test). 2768 J. Neurosci., March 8, 2017 37(10):2764 –2775 Chen et al. Two Types of Episodic Memory Identified with fMRI Table 1. Regions exhibiting greater activity for the life memory test than the Table 3. Regions exhibiting greater activity for the life memory test than the picture memory test (specifically, a contrast of “yes” responses to previously picture memory test (specifically, a contrast of “yes” responses to novel scenes for studied scenes) the life memory test and “yes” responses to studied scenes in the picture memory MNI maximum test) coordinates MNI maximum Size Maximum Region (voxels) z x y z coordinates Size Maximum Region (voxels) z x y z DMN Superior frontal gyrus 76 6.36 ⫺20 31 49 DMN Superior frontal gyrus 47 4.06 ⫺8 12 67 Superior frontal gyrus 79 6.00 ⫺10 63 19 Superior frontal gyrus 33 4.64 22 27 53 Superior frontal gyrus 79 5.65 ⫺20 32 48 Retrosplenial complex 74 5.56 ⫺9 ⫺55 12 Retrosplenial complex 75 4.95 ⫺9 ⫺54 13 Retrosplenial complex 63 4.45 14 ⫺49 14 Retrosplenial complex 68 4.96 10 ⫺53 10 Ant. middle temporal gyrus 65 6.12 62 ⫺6 ⫺19 vMPFC 71 4.84 ⫺4 49 ⫺7 Ant. middle temporal gyrus 59 5.46 ⫺60 ⫺4 ⫺17 dMPFC 46 5.57 ⫺12 50 34 Angular gyrus 76 5.68 48 ⫺68 26 MPFC 29 3.97 8 58 23 Angular gyrus 79 5.50 ⫺45 ⫺67 25 Hippocampus 9 4.41 36 ⫺13 ⫺23 MPFC 72 5.47 ⫺11 60 29 Angular gyrus 74 4.66 ⫺43 ⫺73 27 MPFC 76 5.40 ⫺2 57 7 Superior/middle frontal gyrus 53 4.03 23 27 43 Hippocampus 31 5.17 29 ⫺12 ⫺17 Hand network Hippocampus 47 4.76 ⫺23 ⫺27 ⫺15 Precentral/postcentral gyrus 17 3.75 ⫺9 ⫺32 67 Anterior cingulate gyrus 73 4.51 ⫺1 39 ⫺2 Other Anterior cingulate gyrus 37 4.23 15 40 14 Inferior lateral occipital cortex 51 3.39 47 ⫺69 10 Precuneus cortex 16 3.81 ⫺8 ⫺58 49 Anterior superior/middle temporal gyrus 37 5.34 ⫺58 0 ⫺13 Posterior cingulate cortex 48 4.31 ⫺4 ⫺41 39 Superior frontal gyrus 60 5.11 ⫺13 14 58 Middle frontal gyrus 16 4.22 ⫺36 5 57 Middle temporal gyrus 24 3.53 ⫺45 ⫺52 10 Temporal pole 19 3.92 51 10 ⫺28 Cerebellum Hand network Right IX 31 5.34 1 ⫺51 ⫺49 Precentral gyrus 43 4.18 12 ⫺26 67 Right crus II 27 3.84 22 ⫺80 ⫺36 Precentral gyrus 33 3.96 ⫺11 ⫺31 67 Right crus I 9 3.40 41 ⫺65 ⫺29 Other Network membership was assigned according to the Power et al. (2011) parcellation. Superior frontal gyrus 47 4.06 ⫺8 12 67 Amygdala 32 3.92 ⫺24 ⫺2 ⫺12 Table 4. Regions exhibiting greater activity for the picture memory test than Posterior superior temporal gyrus 30 4.22 ⫺59 ⫺13 ⫺1 the life memory test (specifically, a contrast of “yes” responses to novel scenes Inferior lateral occipital cortex 36 3.70 40 ⫺61 9 for the life memory test and “yes” responses to studied scenes in the picture Inferior frontal gyrus 20 3.55 ⫺46 34 4 memory test) Inferior frontal gyrus 9 3.83 ⫺47 18 25 Cerebellum MNI maximum coordinates Right IX 57 5.13 6 ⫺54 ⫺46 Size Maximum Right crus I 37 4.68 42 ⫺61 ⫺33 Region (voxels) z x y z Right crus II/I 32 4.35 25 ⫺81 ⫺33 FPN subnetwork Network membership was assigned according to the Power et al. (2011) parcellation. Posterior middle IPS 77 5.30 ⫺36 ⫺60 50 Anterior PFC 14 4.67 34 58 ⫺4 Table 2. Regions exhibiting greater activity for the picture memory test than the Anterior middle IPS 81 4.69 44 ⫺49 50 life memory test (specifically, a contrast of “yes” responses to previously studied Middle frontal gyrus 43 4.93 38 14 52 scenes) Middle frontal gyrus 71 4.36 43 20 35 MNI maximum Superior medial PFC 54 4.42 4 36 38 coordinates PMN Size Maximum Region (voxels) z x y z Middle cingulate 60 5.25 2 ⫺29 33 Precuneus cortex 55 5.50 ⫺10 ⫺70 40 FPN subnetwork Precuneus cortex 68 5.60 12 ⫺69 38 Anterior middle IPS 69 4.10 44 ⫺50 53 Frontoparietal Middle frontal gyrus 64 4.23 46 20 34 Middle frontal gyrus 50 4.72 43 43 18 Middle frontal gyrus 10 3.61 43 9 55 Other Superior medial PFC 35 4.54 4 35 42 Orbitofrontal cortex 7 4.68 26 23 ⫺13 Posterior middle IPS 41 3.93 ⫺36 ⫺58 50 Network membership was assigned according to the Power et al. (2011) parcellation. Anterior PFC 15 3.94 34 61 5 PMN Middle cingulate gyrus 60 4.81 1 ⫺29 29 Precuneus cortex 61 4.97 12 ⫺70 39 mates for the “yes” responses to old items for the picture memory test Precuneus cortex 33 4.68 ⫺14 ⫺67 37 and life memory test (i.e., for recognition hits and successful autobi- Frontoparietal ographical retrieval), and obtained the mean of magnitudes for each Middle frontal gyrus 42 4.53 43 45 17 Other task condition for each network or subnetwork. We chose 10 mm Insula 68 5.99 34 24 ⫺5 diameter spheres because numerous studies (Power et al., 2012; Cao et al., 2014; Boly et al., 2015; Geerligs et al., 2015; Thompson and Network membership was assigned according to the Power et al. (2011) parcellation. Fransson, 2015; Lerman-Sinkoff and Barch, 2016) used the same di- ameter for spheres based on the ROIs of Power et al. (2011), and the For reasons that will become clear, we focused on members of the use of 10 mm spheres on our part will facilitate across-experiment default mode network (DMN), the parietal memory network (PMN), comparisons in future research. We then performed paired t tests to and a subnetwork of the frontoparietal network (FPN). We drew 10 determine whether the chosen networks showed differential activities mm spheres around the peaks, obtained the average magnitude esti- for the two types of remembering. It is worth mentioning that the t Chen et al. Two Types of Episodic Memory Identified with fMRI J. Neurosci., March 8, 2017 37(10):2764 –2775 2769 trials (or the reverse order, for half the partici- pants). In both types of tests, half of the items were old, and half were new. For both types of tests, participants made remember/know/new judgments. For autobio- graphical trials, they were instructed to re- spond “remember” when they could recollect specific aspects (time, place, feeling) of an event, “know” when they had a gut feeling of familiarity about the scene (such as having been to a similar place) but could not recollect specific details of an event, and “new” when they were not able to use the scene to recollect an event from their past. For recognition trials, participants were in- structed to answer “remember” when they could recollect something specific about hav- ing seen the scene (e.g., the thought they had or connection they made while seeing it), “know” if they had a gut feeling about having seen the scene but could not remember the specific de- tails, and “new” if they did not remember hav- ing seen the picture in the study phase. Analysis. The primary question for Experi- ment 2 was whether the likelihood of recollec- Figure 4. A contrast of old scenes that elicited “yes” responses in the life memory test to new scenes that elicited “yes” tive remembering might be greater in the life responses in the life memory test. The contrast revealed a sparse map, consisting mostly of members in the parietal memory memory test than in the picture memory test, network. an outcome that would influence interpreta- tion of the results in the fMRI study. As will be Table 5. Regions exhibiting greater activity for studied than nonstudied scenes in seen, this pattern did not occur. the life memory test in successful retrieval trials (trials with “yes” responses) MNI maximum Results Size Maximum coordinates Experiment 1 Region (voxels) z x y z Behavioral results The behavioral performance during the picture memory test and PMN the life memory test is shown in Figure 2. Performance in the Posterior IPL 63 3.8075 ⫺40 ⫺63 47 Posterior IPL 40 4.1245 43 ⫺63 47 picture memory test was quite accurate, with a hit rate of 0.76 and Precuneus cortex 51 3.8358 ⫺6 ⫺68 38 a false alarm rate of 0.08. Accuracy cannot be measured for auto- Precuneus cortex 33 3.8311 13 ⫺64 36 biographical memory, but we can see that participants often Other claimed to be able to retrieve a life memory, whether the cue was Subgenual ACC 19 3.9597 ⫺1 23 ⫺1 an old scene (0.74) or a new scene (0.68). The two percentages do Network membership was assigned according to the Power et al. (2011) parcellation. not differ significantly (t(26) ⫽ 0.754, p ⫽ 0.458). The response times for the test types differed: For hits, the average RT was 1.39 s, whereas for the analogous condition in the tests did not use any “double dipping” because these ROIs were inde- life memory test, the average RT was 2.13 s, (t(26) ⫽ 8.382, p ⫽ pendently defined using coordinates from Power et al., 2011. 7.294 ⫻ 10 ⫺9). For this reason, we entered trial-by-trial RT as a Experiment 2: behavioral experiment covariate in our general linear model of the fMRI data. Both the As will be seen, many of the brain regions more active for the life memory picture memory test and the life memory test had adequate num- test have been implicated previously in recollective remembering (Rugg bers of trials for analysis. Specifically, the number of hit trials for and Vilberg, 2013), in contrast to general feelings of familiarity. One each participant ranged from 9 to 30 (median, 25), and correct possible explanation, therefore, is that the life memory test may elicit rejections ranged from 32 to 64 (median, 57). For the life memory more recollective remembering than the picture memory test (and that it test, yes responses occurred to old scenes with a frequency of 13 to is this difference that drives the differential activity). We test that hypoth- 32 (median, 24), and to new scenes with a frequency of 19 to 60 esis with Experiment 2. (median, 44). Participants. Thirty-eight participants were recruited using Washing- A final observation from the behavioral data is that responses ton University’s subject pool (21 female, ages 18 –21). Informed consent to new scenes differed for the two test types, demonstrating that was obtained from all participants, and the study was conducted in ac- cordance with Washington University’s human research practices. Par- participants understood and followed the instructions (i.e., infre- ticipants either received course credit or $10 per hour. None of the quently claiming to recognize new scenes as having been studied participants in the two experiments overlapped. but frequently reporting being able to use new scenes to trigger Materials. The stimuli consisted of 246 indoor and outdoor scenes col- memories for life events). lected using Google Images, similar to the imaging experiment. Encoding and retrieval instructions and procedures. The procedure of Neuroimaging results the behavioral experiment was very similar to that of the neuroimaging A contrast of old (studied) scenes leading to successful retrieval in the study, except that participants encoded fewer stimuli (96 scenes), all two types of tests revealed differential activity in numerous regions. procedures occurred within the psychology lab, and 96 picture memory The BOLD activity elicited by successful retrieval of previously test trials occurred within a single block, followed by 96 life memory test studied scenes in the picture memory test (i.e., hits) was con- 2770 J. Neurosci., March 8, 2017 37(10):2764 –2775 Chen et al. Two Types of Episodic Memory Identified with fMRI Figure 5. ROIs obtained from the whole-brain analysis align with brain networks identified from an independent data set using graph theory to analyze resting state data (Power et al., 2011). Specifically, the underlays are from a Workbench (Marcus et al., 2011) border file obtained from J. Power (Weill Cornell/NY Presbyterian Hospital, Department of Psychiatry) that was based on the 0.5% tie density modified voxelwise subgraphs. A, The DMN (red underlay) and the ROIs within the DMN emerging from the whole-brain analysis. B, C, PMN and FP subnetwork (tan and blue underlays, respectively), along with the spherical ROIs within those networks. D, Mean activity (and associated SEs) for each of the DMN ROIs in Table 1. Activation for most of these regions (relative to implicit baseline) in the life memory test is evident, as is the tendency for little or no activation during the picture memory test. Conversely, E and F show the reverse tendency for the PMN and FP subnetwork. G–I, Mean activity in the present data set for all regions of the independently defined ROIs from Power et al. (2011). trasted with successful retrieval elicited by previously studied regions sensitive to stimulus repetition, there is very little “contam- scenes in the life memory test (old scenes, for which the subjects’ ination” of recognition signal in the life memory test. responses were “yes, I remember”; Figure 3A, Table 1). Regions Strong hemispheric asymmetry emerged such that most of the more active during autobiographical retrieval included bilateral picture memory ⬎ life memory differences occurred in the right hippocampus, left amygdala, bilateral superior frontal gyrus, an- hemisphere or midline, whereas the opposite pattern (life mem- gular gyrus, medial prefrontal cortex, and bilateral retrosplenial ory ⬎ picture memory) tended to be bilateral but more pro- complex. Regions more activated during recognition hits in- nounced in the left hemisphere. cluded right middle–frontal gyrus, bilateral insula, right inferior Patterns seen in the voxelwise analysis accord with networks parietal cortex, precuneus, and midcingulate cortex (Table 2). defined by resting state fMRI. Regions that exhibited differential Successful retrieval of autobiographical memory for new activity for the two test types were further examined. ROI gener- items relative to recognition hits revealed a similar pattern of ation processes revealed 29 regions differentially activated during results (Fig. 3B, Tables 3, 4). Because of the similarity of the two the two tests (Tables 1, 2). ROIs were categorized based on their contrasts, further comparison between autobiographical retrieval network membership using the parcellation of the Power et al. and recognition refers to the successful autobiographical retrieval (2011) modified voxelwise map. Each of the 18 regions falling given old items versus hits (given the equivalent item history). within the default mode network exhibited greater activity during The contrast involving new items is noteworthy, though, in that the life memory test (Tables 1, 2; Fig. 5). Specifically, Figure 5A this condition is the one typically used in the autobiographical shows the strong correspondence in ROIs from the current data memory literature. set (spheres) overlain on the default mode network (light red The contrast of successful retrieval of autobiographical memory underlay) as identified by Power et al. (2011) based on their for old and new items revealed a sparse map, consisting mainly of voxelwise network parcellation with 0.5% tie density. regions in the recently described parietal memory network (Gilmore Another regularity in the data is that two network communi- et al., 2015; Fig. 4, Table 5). The contrast suggests that aside from ties exhibited the opposite pattern: They were more activated Chen et al. Two Types of Episodic Memory Identified with fMRI J. Neurosci., March 8, 2017 37(10):2764 –2775 2771 during the picture memory test (Fig. 5 B, C). These consisted of Table 6. Regions from Power et al. (2011) used in network analyses the parietal memory network and an unnamed subnetwork of the x y z frontoparietal network (Power et al., 2011; Fig. 4). Regions within DMN both of these networks have been implicated previously in mem- ⫺41 ⫺75 26 ory retrieval (Henson et al., 1999; Yonelinas et al., 2005; Nelson et 6 67 ⫺4 al., 2010; Power et al., 2011). 8 48 ⫺15 The network-level dissociation remains even when using ⫺13 ⫺40 1 ⫺18 63 ⫺9 independently defined network ROIs. To examine whether the dis- ⫺46 ⫺61 21 sociation between successful retrieval in the two memory tests is 43 ⫺72 28 restricted to part of the three networks identified from our con- ⫺44 12 ⫺34 trast or whether the networks as a whole would show the same 46 16 ⫺30 trend, we obtained coordinates for members of the default mode ⫺68 ⫺23 ⫺16 ⫺44 ⫺65 35 network, the PMN, and the FPN subnetwork from the 264 ROIs ⫺39 ⫺75 44 in the Power et al. (2011) parcellation (Table 6). There are 58 ⫺7 ⫺55 27 ROIs in the default mode network defined by Power et al. (2011), 6 ⫺59 35 5 ROIs in the PMN, and 5 ROIs in the FPN subnetwork. ⫺11 ⫺56 16 When averaging over regions within each of the three networks, ⫺3 ⫺49 13 8 ⫺48 31 the results converge with those presented above (Fig. 5G–I). Specif- 15 ⫺63 26 ically, the default mode network exhibited activation in the life ⫺2 ⫺37 44 memory test and little to no activity for the picture memory test, with 11 ⫺54 17 the difference between the two being significant with a two-tailed 52 ⫺59 36 paired t test (t(57) ⫽ 6.511, p ⫽ 2.065 ⫻ 10 ⫺8). Conversely, the two 23 33 48 ⫺10 39 52 memory retrieval networks exhibited activation in the picture mem- ⫺16 29 53 ory test, but little to no activity in the life memory test, with the ⫺35 20 51 difference between the two being statistically significant for the PMN 22 39 39 (t(4) ⫽ 3.608, p ⫽ 0.023), but not reaching significance for the FPN 13 55 38 ⫺10 55 39 subnetwork (t(4) ⫽ 1.710, p ⫽ 0.163). ⫺20 45 39 Similarities in the task activation and manipulation checks. Al- 6 54 16 though the focus of the analyses has been on the differences in 6 64 22 activation and deactivation of the two types of tasks, we show ⫺7 51 ⫺1 (Figs. 6 A, B) that these differences should be understood in light 9 54 3 ⫺3 44 ⫺9 of broad similarities when compared to a low-level baseline. 8 42 ⫺5 These similarities are to be expected in that the tasks were equated ⫺11 45 8 as much as possible and both involved looking at scenes, directing ⫺2 38 36 attention toward the past, making a memory-related decision ⫺3 42 16 (yes/no), and executing a button press; the t tests show activation ⫺20 64 19 ⫺8 48 23 relative to the low-level control of the implicit baseline. Nonethe- 65 ⫺12 ⫺19 less, some of the differences seen in Figure 5 can be gleaned from ⫺56 ⫺13 ⫺10 examining the differences in Figure 6, A and B, the most promi- ⫺58 ⫺30 ⫺4 nent example being in ventromedial frontal cortex, which is ac- 65 ⫺31 ⫺9 tive during the life memory test but strongly deactivated during ⫺68 ⫺41 ⫺5 13 30 59 the picture memory test. 12 36 20 In addition, as a manipulation check and as a way to connect to 52 ⫺2 ⫺16 the prior meta-analysis suggesting pronounced differences (McDer- ⫺26 ⫺40 ⫺8 mott et al. 2009), we consider retrieval success maps for the two types 27 ⫺37 ⫺13 of tests. Figure 6C (blue) shows regions more activated during hits ⫺34 ⫺38 ⫺16 28 ⫺77 ⫺32 than correctly rejected lures for picture memory. This contrast has 52 7 ⫺30 been seen many times in individual studies in the literature (Konishi ⫺53 3 ⫺27 et al., 2000; McDermott et al., 2000) and in meta-analyses (Wagner 47 ⫺50 29 et al., 2005; McDermott et al., 2009; Spaniol et al., 2009; Nelson et al., ⫺49 ⫺42 1 ⫺46 31 ⫺13 2010; Power et al., 2011; Kim, 2013), and shows strong alignment 49 35 ⫺12 with the similar contrast from McDermott et al. (2009) (Fig. 6D); FPN subnetwork (blue in Power et al., 2011, their Fig. 4) regions include dorsal parietal cortex, middle frontal gyrus, insula, 44 ⫺53 47 precuneus, and mid/posterior cingulate cortex. ⫺42 ⫺55 45 The contrast in the present data set most analogous to that typi- 40 18 40 cally used in the autobiographical memory literature is the set of ⫺34 55 4 ⫺42 45 ⫺2 regions more active when novel cues give rise to an autobiographical PMN (salmon in Power et al., 2011, their Fig. 4) memory relative to when they do not (Fig. 6C,D, red). These regions ⫺2 ⫺35 31 in the present data set (Fig. 6C) also align closely with the meta- ⫺7 ⫺71 42 analysis (Fig. 6D): ventral parietal cortex, posterior cingulate cortex, 11 ⫺66 42 4 ⫺48 51 bilateral parahippocampal cortex, and medial frontal cortex. 2 ⫺24 30 Regions showing overlap in these contrasts are depicted in green in Figure 6. Here, we see more extensive overlap than in the 2772 J. Neurosci., March 8, 2017 37(10):2764 –2775 Chen et al. Two Types of Episodic Memory Identified with fMRI Figure 6. A, B, Activity for old items given “yes” responses on the life memory test (A) and the picture memory test (B), relative to an implicit baseline. Despite many differences as a function of what is remembered, one can see many qualitative similarities in the two tests. C, Binarized maps showing “retrieval success” maps for life memory (“yes” responses to new scenes contrasted with “no” responses to new scenes; red) and for picture memory (“yes” responses to old scenes contrasted with “no” responses to new scenes; blue). These maps were created to be similar in construction to those underlying D, a previously reported meta-analysis showing different retrieval networks for autobiographical memory (red), retrieval success on recognition memory (blue), and the overlap of the two networks (green). Data are from McDermott et al. (2009). meta-analysis, possibly reflecting greater power in a within- explained by positing that the life memory test had more recol- subject contrast or features of our autobiographical task that dif- lective processing. The remember/know data suggest that to the fer from those in the literature. Specifically, we see much greater degree that any differences exist, there is more recollective pro- overlap within left frontal cortex, and some along the middle/ cessing reported in the picture memory test. posterior inferior parietal lobule (pIPL), but no overlap in poste- rior cingulate/precuneus. Discussion When participants were asked to retrieve memories in response to Experiment 2 recently encountered scenes, the neural substrates differed depend- The goal of this experiment was to test the hypothesis that the differ- ing on whether the retrieved memory was of having viewed the scene ences observed in Experiment 1 might be attributable to different recently (picture memory test) or whether it was a lifetime event (life reliance on remembering and knowing for the two memory tests. memory test). Regions preferentially active during successful re- For scenes that had been studied, the picture memory test was more trieval in the life memory test fell within the default mode network likely to be accompanied by “remember” experiences than was the (Fig. 5A). Conversely, regions exhibiting greater retrieval-activity in life memory test (Mean ⫽ 48 and 30%, respectively; t(37) ⫽ 5.206, the picture memory test tended to fall within the PMN and a sub- p ⫽ 7.452 ⫻ 10 ⫺6), as seen in Figure 7. In addition, for old items, network of the FPN (Fig. 5B,C). Similar results emerged when non- participants made more “remember” than “know” judgments for studied scenes in the life memory test were contrasted with studied the picture memory test (t(37) ⫽ 5.358, p ⫽ 4.646 ⫻ 10 ⫺6) but not scenes from the picture memory test (Fig. 3B). the life memory test (t(37) ⫽ 0.123, p ⫽ 0.9028). The preponderance In a second experiment, we explored the hypothesis that the of “new” judgments for the nonstudied scenes in the picture mem- life memory test was more likely to invoke vivid recollection than ory test is high, as would be expected; the few false alarms that existed the picture memory test), an idea suggested previously (Cabeza et tended to be “know” responses. al., 2004; Rissman et al., 2016). Specifically, using a remember/ Overall, this pattern of data suggests that the difference be- know procedure (Tulving, 1985; Gardiner, 1988), we found that tween the two types of tests in Experiment 1 cannot be readily contrary to this hypothesis, the picture memory test was more Chen et al. Two Types of Episodic Memory Identified with fMRI J. Neurosci., March 8, 2017 37(10):2764 –2775 2773 memories of more sustained activities (evolving over seconds or minutes, unlike the fleeting presentation of a single word). In addition, the monitoring of the un- folding retrieval episode may differ be- tween the two tests; that is, as the retrieval process develops in the life memory test (and in autobiographical memory), there may be less monitoring for accuracy and temporal specificity (Cabeza et al., 2004), a process sometimes referred to as postre- trieval monitoring (Henson et al., 1999). A similar possibility is that the monitoring component in autobiographical memory comes later in time (Addis et al., 2007; Daselaar et al., 2008), and the short (5 s) trials used in the present experiment may not have allowed enough time for this component to occur. The finding that the picture memory test involved greater activation in compo- nents of the FPN is consistent with these possibilities outlined above regarding grain size, search set size, specificity, and Figure 7. Remember/know/new distributions for the various conditions of Experiment 2. Remember responses (blue) were monitoring (Dobbins et al., 2002, 2003). most common in the picture memory test (for old scenes), not the life memory test. Another possibility is that the life memory test was more recall-like than the picture memory test; although partici- likely to lead to “remember” responses than was the life memory pants made old/new judgments in both retrieval tests, the picture test. Results from Experiment 2 therefore suggest that a straight- memory test was a typical recognition memory test, whereas the forward remember/know account (such that the life memory test life memory test involved cued recall of a lifetime event and then was mostly remembering and the picture memory test mostly a button press, similar to the recognition test. This difference may knowing) cannot readily explain results from Experiment 1. line up well with the aforementioned role of monitoring pro- However, as no remember/know data were collected in Exper- cesses and their influence in recognition. iment 1, further work is necessary before definitive conclusions Furthermore, the differences in ventromedial frontal cortex regarding recollective differences can be drawn. (deactivated in the picture memory test but activated in the life memory test), may be attributable to more prominent self- What are the critical differences between the two types referential processing in the life memory test (Kelley et al., 2002; of tests? Cabeza et al., 2004 St. Jacques et al., 2011). Given the nature of the The two memory tasks were designed with the goal of examining different memory tasks, this should not be surprising, but it nev- memory retrieval of laboratory-learned stimuli and events from ertheless likely contributes to some of the observed difference. one’s lifetime with the memory cues, trial timing and structure, Differences between tests were also observed in the retro- and other details held constant. The data are consistent with the splenial cortex, along with parahippocampal cortex (for a similar hypothesis that episodic retrieval of recently encountered stimuli pattern, see Elman et al., 2013). These regions have been tightly differs from episodic retrieval from one’s life (Roediger and linked with autobiographical memory as well as scene construc- McDermott, 2013). What are the critical features differentiating tion (Hassabis and Maguire, 2007, 2009) and the processing of the two? This question remains to be answered, although we contextual associations (Aminoff et al., 2007; Szpunar et al., 2009; consider some possibilities here. McDermott and Gilmore, 2015). In the current experiment, all of One avenue that may prove fruitful to explore is the grain size the stimuli were scenes, so scene processing was inherent in both of the search space. The picture memory test (and most laboratory tasks. Nevertheless, the greater activity observed in these regions tests) involves a highly constrained search space (i.e., a temporally during the life memory test suggests scene construction may have specific set of stimuli). Conversely, the life memory test (like most been a more prominent component process for this test. autobiographical memory measures) allowed the search of one’s en- Given that the life memory test referred to more remote events tire lifetime up to the point of the beginning of the experiment. than the picture memory test, one might wonder if a difference in Besides the temporal specificity of the search, the absolute number of retention interval contributed to the neural differences. The data, items within the search space was 32 in the picture memory test and however, are difficult to interpret. The system consolidation ac- uncountably large in the life memory test. count posits that as time passes, hippocampal involvement in Temporal specificity of the search space and number of targets memory retrieval decreases (Dudai, 2012; Squire et al., 2015). being searched also aligned with the “episode” being searched for. Although exceptions exist (Rekkas and Constable, 2005), the ma- In the case of the picture memory test, the question was whether jority of memory studies either support or do not contradict this a specific visual stimulus was seen previously, with the picture hypothesis (Niki and Luo, 2002; Takashima et al., 2009; Ya- being a mini episode (Tulving, 1972). The life memory test (and mashita et al., 2009; Söderlund et al., 2012; Furman et al., 2012). autobiographical memory tests in general) tended to invoke In the present study, however, the more remote life memory test 2774 J. Neurosci., March 8, 2017 37(10):2764 –2775 Chen et al. Two Types of Episodic Memory Identified with fMRI produced greater activity in bilateral hippocampus than did the in the literature, engaged many brain regions, especially the more recent picture memory test. One might also wonder if tem- DMN, the PMN, and an FPN subnetwork, differently. This result poral remoteness led to greater activity in ventromedial prefron- supports a dissociation in the processes underlying autobio- tal cortex, a finding reported by many researchers (for review, see graphical memory and laboratory-based memory, while offering Nieuwenhuis and Takashima, 2011). However, given the ventro- a novel paradigm by which different aspects of episodic memory medial prefrontal cortex’s sensitivity to self-relevance (St. Jacques might be explored in future work. et al., 2011) and the presumably higher self-relevance in the life memory condition, the relative contributions cannot be esti- References Addis DR, Wong AT, Schacter DL (2007) Remembering the past and imag- mated within our experimental task. ining the future: Common and distinct neural substrates during event construction and elaboration. Neuropsychologia 45:1363–1377. CrossRef Informing understanding of a recently described parietal Medline memory network Aminoff E, Gronau N, Bar M (2007) The parahippocampal cortex mediates Several of the regions identified as more active for the picture spatial and nonspatial associations. Cereb Cortex 17:1493–1503. CrossRef memory test (specifically, precuneus, midcingulate, and lateral Medline Boly M, Sasai S, Gosseries O, Oizumi M, Casali A, Massimini M, Tononi G parietal cortex) correspond to a sparse network recently intro- (2015) Stimulus set meaningfulness and neurophysiological differentia- duced as the PMN (Gilmore et al., 2015). This set of regions was tion: A functional magnetic resonance imaging study. PLoS One 10. shown to be a functional network as identified by functional Boynton GM, Engel SA, Glover GH, Heeger DJ (1996) Linear systems anal- connectivity analyses of resting state data (Power et al., 2011; Yeo ysis of functional magnetic resonance imaging in human V1. J Neurosci et al., 2011) and was hypothesized to reflect the perceived novelty 16:4207– 4221. Medline or familiarity of a particular stimulus. The finding within the Cabeza R, Prince SE, Daselaar SM, Greenberg DL, Budde M, Dolcos F, LaBar current data set that the network activates more to old stimuli KS, Rubin DC (2004) Brain activity during episodic retrieval of autobi- ographical and laboratory events: An fMRI study using a novel photo whose oldness is salient or task relevant (relative to old stimuli paradigm. J Cogn Neurosci 16:1583–1594. CrossRef Medline whose oldness is irrelevant to the task) is consistent with the Cao H, Plichta MM, Schäfer A, Haddad L, Grimm O, Schneider M, Esslinger hypothesis forwarded by Gilmore et al. (2015) (Fig. 3) and sug- C, Kirsch P, Meyer-Lindenberg A, Tost H (2014) Test-retest reliability gests a role for attention to the familiarity in driving activity of the of fMRI-based graph theoretical properties during working memory, network. These data are also broadly consistent with prior find- emotion processing, and resting state. Neuroimage 84:888 –900. CrossRef ings that certain regions associated with the PMN—notably, the Medline Cohen J (1993) PsyScope: An interactive graphic system for designing and left pIPL—show differences in recognition activity depending on controlling experiments in the psychology laboratory using Macintosh whether one is expecting a novel or familiar item at test (O’Connor et computers. Behav Res Methods Inst Comput 25:257–271. CrossRef al., 2010; Jaeger et al., 2013) and whether the mnemonic informa- Conway MA, Turk DJ, Miller SL, Logan J, Nebes RD, Meltzer CC, Becker JT tion is task relevant at test (Elman and Shimamura, 2011; Rosen (1999) A positron emission tomography (PET) study of autobiographi- et al., 2016). cal memory retrieval. Memory 7:679 –702. CrossRef Medline Daselaar SM, Rice HJ, Greenberg DL, Cabeza R, LaBar KS, Rubin DC (2008) The spatiotemporal dynamics of autobiographical memory: Neural cor- Generalization: are there multiple kinds of episodic memory? relates of recall, emotional intensity, and reliving. Cereb Cortex 18:217– We return now to the question posed in the title of our paper. 229. CrossRef Medline Although no single fMRI experiment can definitively answer Dobbins IG, Foley H, Schacter DL, Wagner AD (2002) Executive control such a fundamental question, the present results clearly demon- during episodic retrieval: Multiple prefrontal processes subserve source strate that the type of “episode” influences the neural substrates memory. Neuron 35:989 –996. CrossRef Medline of episodic memory. Specifically, distinct functional systems can Dobbins IG, Rice HJ, Wagner AD, Schacter DL (2003) Memory orientation be recruited, depending on the episodes being retrieved. A prac- and success: Separable neurocognitive components underlying episodic recognition. Neuropsychologia 41:318 –333. CrossRef Medline tical implication of this finding is that the two traditions of study- Dudai Y (2012) The restless engram: consolidations never end. Ann Rev ing episodic memory (autobiographical and laboratory) invoke a Neurosci 35:227–247. CrossRef Medline common name (episodic memory) for different collections of Elman JA, Shimamura AP (2011) Task relevance modulates successful re- processes. Future work that uses different forms of “laboratory” trieval effects during explicit and implicit memory tests. Neuroimage 56: and “autobiographical memory” test will be crucial to under- 345–353. CrossRef Medline standing the importance and generality of this conclusion [al- Elman JA, Cohn-Sheehy BI, Shimamura AP (2013) Dissociable parietal re- gions facilitate successful retrieval of recently learned and personally fa- though a meta-analysis by McDermott et al. (2009) suggests that miliar information. Neuropsychologia 51:573–583. CrossRef Medline we will see similar effects in other situations as well]. Furman O, Mendelsohn A, Dudai Y (2012) The episodic engram trans- Furthermore, it is important to acknowledge that the term formed: Time reduces retrieval-related brain activity but correlates with “episodic memory” has morphed considerably over the years, memory accuracy. Learn Mem 19:575–587. CrossRef Medline becoming increasingly complex and theory bound (see Szpunar Gardiner JM (1988) Functional aspects of recollective experience. Mem and McDermott, 2008). Here we adopt the earliest definition, Cogn 16:309 –313. CrossRef regarding memory for events or episodes, where an event can be Geerligs L, Renken RJ, Saliasi E, Maurits NM, Lorist MM (2015) A brain- wide study of age-related changes in functional connectivity. Cereb Cor- as small as a word in a list (Tulving, 1972). The present data cast tex 25:1987–1999. CrossRef Medline doubt on the assumption that words on a list are “mini episodes” Gilmore AW, Nelson SM, McDermott KB (2015) A parietal memory net- in a way that is usefully comparable to life episodes, but we must work revealed by multiple MRI methods. Trends Cogn Sci 19:534 –543. stress that the degree to which the differences could be minimized CrossRef Medline with increasingly similar tasks (e.g., equating search set size or Hassabis D, Maguire EA (2007) Deconstructing episodic memory with con- self-referential processing) remains to be addressed. struction. Trends Cogn Sci 11:299 –306. CrossRef Medline Hassabis D, Maguire EA (2009) The construction system of the brain. Philos Trans R Soc Lond B Biol Sci 364:1263–1271. CrossRef Medline Conclusion Hassabis D, Kumaran D, Maguire EA (2007) Using imagination to under- In summary, this study demonstrated that laboratory-based and stand the neural basis of episodic memory. J Neurosci 27:14365–14374. autobiographical retrieval, assessed using methods typically used CrossRef Medline Chen et al. Two Types of Episodic Memory Identified with fMRI J. Neurosci., March 8, 2017 37(10):2764 –2775 2775 Henson RN, Rugg MD, Shallice T, Josephs O, Dolan RJ (1999) Recollection retrieval does not become independent of the hippocampus: An fMRI and familiarity in recognition memory: an event-related functional mag- study contrasting very recent with remote events. J Cogn Neurosci 17: netic resonance imaging study. J Neurosci 19:3962–3972. Medline 1950 –1961. CrossRef Medline Jaeger A, Konkel A, Dobbins IG (2013) Unexpected novelty and familiarity Rissman J, Chow TE, Reggente N, Wagner AD (2016) Decoding fMRI sig- orienting responses in lateral parietal cortex during recognition judg- natures of real-world autobiographical memory retrieval. J Cogn Neuro- ment. Neuropsychologia 51:1061–1076. CrossRef Medline sci 28:604 – 620. CrossRef Medline Kelley WM, Macrae CN, Wyland CL, Caglar S, Inati S, Heatherton TF (2002) Roediger HL 3rd, McDermott KB (2013) Two types of event memory. Proc Finding the self? An event-related fMRI study. J Cogn Neurosci 14:785– Natl Acad Sci U S A 110:20856 –20857. CrossRef Medline 794. CrossRef Medline Rosen ML, Stern CE, Michalka SW, Devaney KJ, Somers DC (2016) Cogni- Kim H (2013) Differential neural activity in the recognition of old versus tive control network contributions to memory-guided visual attention. new events: An activation likelihood estimation meta-analysis. Hum Cereb Cortex 26:2059 –2073. Medline Brain Mapp 34:814 – 836. CrossRef Medline Rugg MD, Vilberg KL (2013) Brain networks underlying episodic memory Konishi S, Wheeler ME, Donaldson DI, Buckner RL (2000) Neural corre- retrieval. Curr Opin Neurobiol 23:255–260. CrossRef Medline lates of episodic retrieval success. Neuroimage 12:276 –286. CrossRef Snyder AZ (1996) Difference image vs. ratio image error function forms in Medline PET-PET realignment. In: Quantification of brain function using PET Konkle T, Brady TF, Alvarez GA, Oliva A (2010) Scene memory is more (Myer R, Cunningham VJ, Bailey DL, Jones T, eds), pp 131–137. San detailed than you think: The role of categories in visual long-term mem- Diego: Academic. ory. Psychol Sci 21:1551–1556. CrossRef Medline Söderlund H, Moscovitch M, Kumar N, Mandic M, Levine B (2012) As time Lancaster JL, Glass TG, Lankipalli BR, Downs H, Mayberg H, Fox PT (1995) goes by: hippocampal connectivity changes with remoteness of autobio- A modality-independent approach to spatial normalization of tomo- graphical memory retrieval. Hippocampus 22:670 – 679. Medline graphic images of the human brain. Hum Brain Mapp 3:209 –223. Spaniol J, Davidson PS, Kim AS, Han H, Moscovitch M, Grady CL (2009) CrossRef Event-related fMRI studies of episodic encoding and retrieval: Meta- Lerman-Sinkoff DB, Barch DM (2016) Network community structure al- analyses using activation likelihood estimation. Neuropsychologia 47: terations in adult schizophrenia: identification and localization of altera- 1765–1779. CrossRef Medline tions. Neuroimage 10:96 –106. CrossRef Medline Squire LR, Genzel L, Wixted JT, Morris RG (2015) Memory consolidation. Marcus DS, Harwell J, Olsen T, Hodge M, Glasser MF, Prior F, Jenkinson M, Cold Spring Har Perspect Biol 7:a021766. CrossRef Laumann T, Curtiss SW, Van Essen DC (2011) Informatics and data St. Jacques PL, Conway MA, Lowder MW, Cabeza R (2011) Watching my mining tools and strategies for the human connectome project. Front mind unfold versus yours: An fMRI study using a novel camera technol- Neuroinform 5:4. Medline ogy to examine neural differences in self-projection of self versus other McAvoy MP, Ollinger JM, Buckner RL (2001) Cluster size thresholds for perspectives. J Cogn Neurosci 23:1275–1284. CrossRef Medline assessment of significant activation in fMRI. Neuroimage 6:198. CrossRef Summerfield JJ, Hassabis D, Maguire EA (2009) Cortical midline involve- McDermott KB, Gilmore AW (2015) The role of context in understanding ment in autobiographical memory. Neuroimage 44:1188 –1200. CrossRef similarities and differences in remembering and episodic future thinking. Medline In: The psychology of learning and motivation (Ross BH, ed), pp 45–76. Szpunar KK, McDermott KB (2008) Episodic memory: an evolving con- Elsevier, Cambridge, MA.