|
|
| Line 1: |
Line 1: |
| | __NOTOC__ | | __NOTOC__ |
| − | Test link to anchor on pubs page: [[Publications#AddiKaha04b|Addis & Kahana, 2004]]
| + | The Computational Memory Lab uses mathematical modeling and computational techniques to study human memory. We apply these quantitative methods both to data from laboratory studies of human memory and from electrophysiological studies done on patients with implanted electrodes. |
| | | | |
| − | = New site: thoughts =
| + | Our current research is focused on basic mechanisms of episodic, spatial, and working memory. |
| − | * Sidebar for news feed of upcoming and recent lab events. Could remain on every page (available in some themes: http://moinmo.in/ThemeArchiveMarket).
| + | |
| − | ** Pull events from a CML .ics calendar? Or just make it a Google Apps calendar? (http://www.google.com/a/help/intl/en/admins/editions.html)
| + | |
| − | ** testing calendar on the web (via Google Apps?)
| + | |
| − | * New logo?
| + | |
| − | * ''Participate in research'' page (we should make a participate@psych email alias or something)
| + | |
| | | | |
| − | = Things to explain on the wiki =
| + | We are located at: |
| − | * How to link to files for the external world
| + | 3401 Walnut St. (Entrance C), Room 303 |
| − | * How to <code><nowiki>scp</nowiki></code> files to the server
| + | Philadelphia, PA 19104 |
| | + | Tel. 215-746-3500; Fax 215-746-6848 |
| | + | For directions to the lab, click here. (pdf) |
| | | | |
| − | = Introduction to the Main Research Areas = | + | = Episodic Memory = |
| − | Research in the CML revolves around the study of human memory, combining approaches from traditional experimental psychology, computational modeling, and experimental neuroscience. The laboratory study of human memory began more than a century ago, and although many behavioral phenomena have been discovered and well characterized, the development of predictive quantitative theories is still in its infancy. Our research aims to develop and test theories that address both behavioral and physiological data on human memory function.
| + | |
| | | | |
| − | Much of our recent work aims to understand the physiological underpinnings of human memory function. This work uses both invasive and non-invasive brain recordings. Invasive recordings (intracranial EEG, or iEEG) are taken from patients undergoing treatment for epilepsy or brain tumors to study the physiological correlates of memory function. Because iEEG recordings are taken for clinical purposes, and there is minimal risk associated with performing cognitive tasks, this approach provides a unique opportunity to study the neurobiology of human cognitive function. Non-invasive recordings of electrical activity at the scalp (scalp EEG) provides an excellent compliment to our iEEG research. Whereas intracranial EEG can only be recorded from clinical populations who may exhibit slightly different brain function, scalp EEG can be recorded from normal, healthy young adults. By comparing the highly localized, artifact-free intracranial potentials with those obtained at the scalp, we are able to learn more about the electrophysiological correlates of memory than we would using either method alone.
| + | Episodic memory refers to memory for events that are embedded in a temporal context. This includes both memory for significant life events and memory for common daily activities. In the laboratory, episodic memory is investigated by presenting lists of words for study, and then asking participants to recall the words. We have focused on the use of conditional probability and latency analysis (Kahana, M. J., 1996) to examine how participants transition from one recalled word to the next. These techniques quantify the order in which participants recall list items and the inter-response times between successive recalls (see Fig. 1). |
| | | | |
| − | The following three pages describe research being conducted in the CML within three specific problem domains: | + | Fig. 1: The contiguity effect in free recall. This curve shows the probability of making a recall to serial position i+lag immediately following recall of serial position i---that is, the conditional-response probability (CRP) as a function of lag. |
| − | # ["Mechanisms of Episodic Memory"],
| + | |
| − | # ["Navigational Spatial Memory"], and
| + | |
| − | # ["Recognition Memory"].
| + | |
| | | | |
| − | Given the complexity of human memory, its reliance on a variety of brain structures and mechanisms, and its relevance to so many distinct, yet interrelated, facets of human experience, a multipronged approach to its study seems most profitable. Such an approach allows insights and methods developed within one domain to inform the others.
| + | Fig. 2: Brain oscillations associated with successful encoding are reinstated during correct retrieva. The top row of brain maps contrasts gamma-band oscillatory activity during the two second item presentation for items subsequently recalled and those that were forgotten. The bottom row contrasts gamma-band oscillations during the 500 milliseconds preceding recall verbalization for correct items and for prior-list intrusions. In each map, red corresponds to regions where the contrast was significant, gray to non-significant contrasts, and black indicates brain regions excluded from the analysis due to insufficient electrode coverage. |
| | | | |
| − | = Current Research Projects =
| + | To explain the recency and contiguity effects in free recall, Howard and Kahana (2002) developed the Temporal Context Model of episodic memory. TCM is a distributed memory model that specifies the mechanisms of contextual drift and contextual retrieval. Through the drift mechanism, TCM describes how a temporal code is created by the integration of recently retrieved contextual states. As such, TCM represents the first formal model of how memories become 'episodic' (linked to the time when they occurred). TCM also provides an alternative explanation for associative tendencies in recall. Rather than resulting from co-occurrence in short-term memory (the standard earlier view), TCM suggests that these tendencies appear because recall of an item recovers the temporal context for the item, which in turn cues recall of subsequent items. Similarly, recency effects appear because the temporal context at the time of the memory test is most similar to the temporal context associated with recent items. Unlike short-term memory based models, TCM predicts that recency and associative effects should be approximately time-scale invariant (Howard, M. W. and Kahana, M. J., 1999, Sederberg, et al., 2008). |
| − | The following sections outline the major projects that are ongoing in the CML. A listing of recent publications, along with PDFs of the papers and raw data, can be found on the lab website. Major effort is currently directed toward the study of verbal episodic memory, as measured using free recall, serial recall and item recognition tasks, as well as both verbal and visual working memory, and spatial navigation. To increase our understanding of these aspects of human memory we employ a combination of electrophysiological, computational, and behavioral methods. The electrophysiological methods include iEEG recordings from neurosurgical patients and scalp EEG recordings from healthy young adults. For our scalp-recording studies at UPenn, we are using a 128 channel EGI system. Currently, our intracranial studies are being carried out at the following hospitals: Children's Hospital, Boston, Brigham & Woman's Hospital, Boston, Freiburg Epilepsy Center in Germany, UCLA medical center in Los Angeles, Children's Hospital of Philadelphia, the Hospital of the University of Pennsylvania (HUP) and the Mayo Clinic in Rochester, MN. At UCLA and the Mayo clinic we record both single cell responses and local-field potentials; at all of the other hospitals we record local field potentials. We are also setup to record single cell responses at HUP when the next appropriate patient becomes available.
| + | |
| | | | |
| − | == iEEG and Scalp EEG Projects ==
| + | In addition to behavioral and theoretical analyses of episodic memory, we also explore the neurophysiology of episodic memory with both scalp and intracranial electroencephalographic (iEEG) recordings. Intracranial recordings can be obtained from epilepsy patients who have had electrodes surgically implanted on the cortical surface of the brain or through the medial temporal lobes (including hippocampus) as part of the clinical process of localizing seizure foci. One focus of this research is to determine the oscillatory correlates of successful episodic memory formation. Our lab has found that 44-100 Hz (gamma) brain oscillations increase while participants are studying words that they will successfully, as opposed to unsuccessfully, recall (Sederberg, et al., 2006). The same distribution of gamma activity across both hippocampus and neocortex is reactivated just prior to recalling an item, with higher levels of gamma predicting whether or not the recalled item was actually studied (Sederberg, et al., 2007; see Fig. 2). |
| − | Three main experimental paradigms are currently being run with both iEEG and scalp EEG recordings. Free Recall involves studying a list of words and then, after a brief filled delay, trying to recall them in any order. [[YellowCab]] II is our spatial navigation game, whereby we can study the behavioral and electrophysiological correlates of memory during spatial navigation. Finally, multi-modal Sternberg (MMS) involves the presentation of short lists consisting of verbal or non-verbal stimuli (e.g., words, letters, faces, or sinusoidal gratings); subjects then make recognition judgments on these lists, enabling us to compare working memory for both verbal and non-verbal stimuli. In addition to the three currently active paradigms, we have also amassed a large dataset on the Sternberg task with consonants as stimuli, and a small dataset on paired associate memory. The following is a brief description of each of the three main paradigms.
| + | |
| − | | + | |
| − | === Free Recall ===
| + | |
| − | Free recall is the standard test of episodic memory. Subjects study a list of words and then try to recall as many words as they can in any order. Analyzing iEEG data, Sederberg et al (2003,2006) found that and gamma (> 32-80 Hz) oscillations increase, while alpha/beta (8-20 Hz) oscillations decrease during the encoding of a word that is later recalled. The increase in gamma oscillations during successful encoding is primarily seen in two brain regions: the left inferior prefrontal cortex and the lateral temporal cortex (Sederberg et al., 2006). In an experiment investigating temporal/semantic interactions as a guide for recall, Sederberg et al. (in preparation) have shown that ERP activity at time of encoding can predict whether items will later be recalled by way of temporal and/or semantic associations. Largest ERP increases are seen in the frontal regions when forming an association between nearby and semantically similar words, as well as in the posterior regions when forming an association between distant and semantically similar words. "Followup free recall studies are begin carried out by Dr. Sean Polyn, Ilana Jerud, Dov Kogen, Vadim Koshkin, Neal Morton"
| + | |
| − | | + | |
| − | === [[YellowCab]] II ===
| + | |
| − | Building on previous work by Dr. Jeremy Caplan, Dr. Arne Ekstrom, Ehren Newman, and Igor Korolev we are continuing our studies of spatial navigation using both iEEG and scalp EEG recording methods (for more information, see section 2.2). Josh Jacobs is analyzing [[YellowCab]] data to better understand the relation of cellular activity and brain oscillations during navigation. Matt Mollison and Igor Korolev are analyzing scalp EEG data collected during [[YellowCab]]. Dr. Christoph Weidemann is now beginning new analyses of our Yellow Cab datatsets. Jonathan Miller is working with Dr. Sean Polyn on a project designed to assess the role of temporal and spatial factors in memory retrieval using a new variant of the Yellow Cab task. We continue to work with Arne Ekstrom and Itzhak Fried at UCLA, where we are collecting single-unit recordings and local field potentials from patients with epilepsy. "Dr. Sean Polyn, Dr. Christoph Weidemann, Josh Jacobs, Matt Mollison, Jonathan Miller, Jeremy Manning"
| + | |
| − | | + | |
| − | === Visual and Spatial Working Memory ===
| + | |
| − | Previous work in the lab has investigated the role of brain oscillations in short term memory for linguistic materials, such as letters (Raghavachari et al., 2001,2006; Rizzuto et al. 2003, 2005; Jacobs et al, 2006), and more recently, comparing the oscillatory correlates of working memory for verbal and nonverbal materials (see Hwang et al, 2005). Marieke van Vugt is furthering this line of research by examining the physiological correlates of summed-similarity and proactive interference in working memory for verbal and nonverbal stimuli, using both scalp EEG and iEEG data. Marieke's work is building on the NEMO model of Kahana and Sekuler (2002).
| + | |
| − | | + | |
| − | === Phase Locking ===
| + | |
| − | | + | |
| − | In this project, we examine whether human brain oscillations function as a broad neural timing signal. We find that the spiking of ''phase-locked'' neurons occurs at particular phases of neuronal oscillations at various frequencies. Phase-locked neurons activate at various phases of theta and delta oscillations (1-10 Hz), but spike primarily at the trough of gamma oscillations (30-100 Hz). These findings show that brain oscillations facilitate precise spike timing in humans, and suggest that gamma and theta oscillations play complementary roles in this system.
| + | |
| − | | + | |
| − | attachment:phaseLocking.png
| + | |
| − | | + | |
| − | == Modeling Projects ==
| + | |
| − | === Temporal Context Model (TCM) ===
| + | |
| − | Building on the original model proposed by Howard & Kahana (2002), Dr. Per Sederberg has developed a new version of TCM (called TCM-A) that uses a more realistic retrieval process to fit a variety of data on free recall, including data on response latencies and amnesia. Sederberg's work is currently being written up for publication.
| + | |
| − | | + | |
| − | === Strength Based Memory Models ===
| + | |
| − | We have recently developed an extension of the SAM model called eSAM, that accounts for the role of semantic and pre-experimental information on episodic memory (see Sirotin, Kimball, & Kahana, 2005). This work is a collaboration with Dr. Dan Kimball, assistant professor of psychology at the University of Texas, Arlington, and Gene Sirotin, a former undergraduate who is pursuing a Ph.D. in Neuroscience at Columbia. Our latest work uses eSAM to account for data on the false memory (DRM) effect (Kimball et al., in revision).
| + | |
| − | | + | |
| − | === Noisy Exemplar Model (NEMO) ===
| + | |
| − | NEMO was presented initially in Kahana and Sekuler (2002), and follow up work has been carried out by Nosofsky (2005), Kahana et al. (in press) and Yotsumoto (in press). Marieke is working on extending NEMO to account for inter-list interference effects (proactive interference). In addition, she works on trying to understand how NEMO could inform our analysis of brain oscillations during this task.
| + | |
| − | | + | |
| − | === Magellan: An Ideal Navigator Model ===
| + | |
| − | Manning, Kahana & Sekuler (submitted) proposed a model of spatial navigation in Yellow Cab and showed how it fit learning curves and environment difficulty in the studies reported by Newman et al. (2006) and Korolev and Mollison (in preparation).
| + | |
The Computational Memory Lab uses mathematical modeling and computational techniques to study human memory. We apply these quantitative methods both to data from laboratory studies of human memory and from electrophysiological studies done on patients with implanted electrodes.
Our current research is focused on basic mechanisms of episodic, spatial, and working memory.
We are located at:
3401 Walnut St. (Entrance C), Room 303
Philadelphia, PA 19104
Tel. 215-746-3500; Fax 215-746-6848
For directions to the lab, click here. (pdf)
Episodic memory refers to memory for events that are embedded in a temporal context. This includes both memory for significant life events and memory for common daily activities. In the laboratory, episodic memory is investigated by presenting lists of words for study, and then asking participants to recall the words. We have focused on the use of conditional probability and latency analysis (Kahana, M. J., 1996) to examine how participants transition from one recalled word to the next. These techniques quantify the order in which participants recall list items and the inter-response times between successive recalls (see Fig. 1).
Fig. 1: The contiguity effect in free recall. This curve shows the probability of making a recall to serial position i+lag immediately following recall of serial position i---that is, the conditional-response probability (CRP) as a function of lag.
Fig. 2: Brain oscillations associated with successful encoding are reinstated during correct retrieva. The top row of brain maps contrasts gamma-band oscillatory activity during the two second item presentation for items subsequently recalled and those that were forgotten. The bottom row contrasts gamma-band oscillations during the 500 milliseconds preceding recall verbalization for correct items and for prior-list intrusions. In each map, red corresponds to regions where the contrast was significant, gray to non-significant contrasts, and black indicates brain regions excluded from the analysis due to insufficient electrode coverage.
To explain the recency and contiguity effects in free recall, Howard and Kahana (2002) developed the Temporal Context Model of episodic memory. TCM is a distributed memory model that specifies the mechanisms of contextual drift and contextual retrieval. Through the drift mechanism, TCM describes how a temporal code is created by the integration of recently retrieved contextual states. As such, TCM represents the first formal model of how memories become 'episodic' (linked to the time when they occurred). TCM also provides an alternative explanation for associative tendencies in recall. Rather than resulting from co-occurrence in short-term memory (the standard earlier view), TCM suggests that these tendencies appear because recall of an item recovers the temporal context for the item, which in turn cues recall of subsequent items. Similarly, recency effects appear because the temporal context at the time of the memory test is most similar to the temporal context associated with recent items. Unlike short-term memory based models, TCM predicts that recency and associative effects should be approximately time-scale invariant (Howard, M. W. and Kahana, M. J., 1999, Sederberg, et al., 2008).
In addition to behavioral and theoretical analyses of episodic memory, we also explore the neurophysiology of episodic memory with both scalp and intracranial electroencephalographic (iEEG) recordings. Intracranial recordings can be obtained from epilepsy patients who have had electrodes surgically implanted on the cortical surface of the brain or through the medial temporal lobes (including hippocampus) as part of the clinical process of localizing seizure foci. One focus of this research is to determine the oscillatory correlates of successful episodic memory formation. Our lab has found that 44-100 Hz (gamma) brain oscillations increase while participants are studying words that they will successfully, as opposed to unsuccessfully, recall (Sederberg, et al., 2006). The same distribution of gamma activity across both hippocampus and neocortex is reactivated just prior to recalling an item, with higher levels of gamma predicting whether or not the recalled item was actually studied (Sederberg, et al., 2007; see Fig. 2).
