Arne D. Ekstrom - US grants

[unreadable] DESCRIPTION (provided by applicant): The goal of this proposal is to correlate the hemodynamic BOLD signal with cellular electrophysiological activity in the human medial temporal lobes. This project proposes to perform high-resolution structural imaging studies of the hippocampal area during performance of navigation tasks. Using computational methods that allow unfolding of activity in subregions of the brain, I will examine fMRI BOLD activity in the subnuclei of the hippocampus and specific layers of the parahippoccampal region. These same patients will subsequently be implanted with depth electrodes in these brain regions as part of resection planning for treatment of pharmacologically intractable epilepsy. Using an 128 channel recording system that allows continuous sampling of wide band electrophysiological data, both EEG and single neuron responses in the hippocampal area will be examined. By examining EEG and single neuron responses to behaviorally relevant events, these signals will then be correlated with the BOLD signal. Testing with control subjects will provide insight into which activations in patients may represent impairments due to seizures and can justify comparing our findings on correlations between BOLD and electrophysiology to normal brain activity. [unreadable] [unreadable]

DESCRIPTION (provided by applicant): The goal of this project is to determine the neural basis of human episodic memory using an innovative combination of high-resolution functional magnetic resonance imaging (fMRI) and intracranial EEG (iEEG). Episodic memory involves knowing where and when an event occurred relative to other events, both of which depend critically on the hippocampus. Yet exactly how and in what manner the hippocampus codes spatial and temporal aspects of episodic memory remains unclear and understudied, particularly in humans where the primary focus of research has been on verbal episodic memory. Using a newly developed experimental paradigm from our lab, we will study the spatial and temporal components of episodic memory, focusing on two critical processes underlying these components: representation and binding. To map these onto the hippocampal circuit, we employ high-resolution fMRI. In contrast to some previous models developed in the rodent, we hypothesize that spatio-temporal representation is a function shared across hippocampal subregions but that subregion CA3/DG plays a distinct role in parsing elements of context to represent an episode. We further hypothesize a central role for hippocampal subregion CA1 in spatio-temporal binding based on its unique connectivity, in contrast to previous models that have focused on CA3/DG. Collaborating with a team of neurologists and neurosurgeons at two different hospitals, we will also employ iEEG in patients undergoing seizure monitoring. This complementary approach will allow us to determine a separate yet critical component of episodic memory: how does coordinated neural activity in the hippocampus, long linked with spatial navigation in the rodent but understudied in humans, underlie representation and binding of spatio-temporal memory? Overall, our proposed experiments will provide novel insight into the neural basis of episodic memory as they take a new approach to this issue paradigmatically and methodologically and allow us to test several different models of hippocampal function, including our model. Neurodegenerative diseases such as stroke, epilepsy, and schizophrenia impact hippocampal subregion function and coordinated neural activity there, often resulting in symptoms of spatial disorientation and temporal confusion in patients afflicted with these conditions. Our approach thus will also have significant implications for neural diseases that affect the hippocampus.

DESCRIPTION (provided by applicant): The goal of this project is to determine whether low frequency oscillations serve as a mechanism for coordinating cortical areas underlying human episodic memory. To address this issue, we will first employ innovative multilobular electrocortigraphic (ECOG) recordings in patients to determine the sets of interconnected brain areas underlying episodic memory, which previous research and our preliminary work strongly suggest to include the medial temporal lobes, prefrontal cortex, and parietal cortex. We will then perturb areas of high connectivity (hubs) in the same patients using chronometric stimulation, which involves simultaneous recording and stimulation from two different brain regions. Chronometric stimulation is advantageous because it involves stimulation that mimics the frequency and amplitude of on-going recorded activity in another brain region, potentially mitigating unwanted spread of stimulation to other brain areas and at the same time providing insight into how neural communication might actually occur. We will employ two different stimulation methods with this approach, either in phase or out of phase stimulation with the on-going recorded oscillations in connecting hubs. This will allow us to determine whether 1) areas with high degrees of connectivity (hubs) are necessary for episodic memory 2) whether in- phase, coherent oscillatory can enhance episodic memory retrieval 3) whether out-of- phase oscillations result in decrements in memory performance. Our approach here combines innovative tools, such as electrocorticography and chronometric stimulation in humans and analysis techniques involving graph theory. These in turn will allow us to advance our understanding of how and in what manner networks of brain regions interact as part of their role in episodic memory. This work is relevant to clinical research because it can provide insight into the extent to which other brain regions can compensate for lost function following stroke-related lesions to the medial temporal lobes, a known hub in episodic memory. It will also advance our understanding of potential ways to design and implement deep brain stimulators to treat cognitive impairments accompanying neural disease. For example, if the experiments outlined here are successful, they would imply that devices that time stimulation to be in-phase with distant recorded oscillatory activity could restore or even enhance impaired memory function in patients suffering from neural disease.

? DESCRIPTION (provided by applicant): The goal of this project is to test a novel theoretical framework that relates memory and navigation functions of the human medial temporal lobes. While numerous studies in rats and humans suggest the importance of the medial temporal lobes to navigation and memory, these two functions remains poorly linked. Our theoretical framework argues that damage to the human medial temporal lobes results in deficits in both high-resolution perception and binding of details in memory. Thus, one prediction of this model is that patients with medial temporal lobe damage will show deficits in encoding high-resolution spatial details during navigation. This will manifest specifically in impairments in spatial precison during navigation rather than whether patients use distal landmarks (allocentric) or self-orientation (egocentric) cues to navigate. This contrasts with other models of navigation that suggest that MTL damage will impair allocentric but not egocentric navigation. We will test our model by having patients navigate a large arena, searching for a hidden platform, and compare their performance with a group of age and IQ matched healthy controls. A second prediction of our model is that medial temporal lobe damage will impair binding of spatiotemporal details during navigation. We will test this prediction of our model by comparing patient performance when they must remember multiple spatial locations over trials. We expect patient performance to worsen as a function of the number of locations they must remember compared to controls. This contrasts with the predictions of other models of navigation and memory, which suggest impairments in allocentric memory regardless of the number of locations the patient must remember. Our model thus provides novel yet testable predictions that this R03 seed grant will be instrumental in helping us to develop. Use of innovative tools such as virtual reality and high-resolution magnetic resonance imaging (MRI) targeting the medial temporal lobes will also further our understanding of how neural disease such as stroke impacts both brain function and cognition more generally.

DESCRIPTION (provided by applicant): The goal of this project is to determine the neural basis of human episodic memory using an innovative combination of high-resolution functional magnetic resonance imaging (fMRI) and intracranial EEG (iEEG). Episodic memory involves knowing where and when an event occurred relative to other events, both of which depend critically on the hippocampus. Yet exactly how and in what manner the hippocampus codes spatial and temporal aspects of episodic memory remains unclear and understudied, particularly in humans where the primary focus of research has been on verbal episodic memory. Using a newly developed experimental paradigm from our lab, we will study the spatial and temporal components of episodic memory, focusing on two critical processes underlying these components: representation and binding. To map these onto the hippocampal circuit, we employ high-resolution fMRI. In contrast to some previous models developed in the rodent, we hypothesize that spatio-temporal representation is a function shared across hippocampal subregions but that subregion CA3/DG plays a distinct role in parsing elements of context to represent an episode. We further hypothesize a central role for hippocampal subregion CA1 in spatio-temporal binding based on its unique connectivity, in contrast to previous models that have focused on CA3/DG. Collaborating with a team of neurologists and neurosurgeons at two different hospitals, we will also employ iEEG in patients undergoing seizure monitoring. This complementary approach will allow us to determine a separate yet critical component of episodic memory: how does coordinated neural activity in the hippocampus, long linked with spatial navigation in the rodent but understudied in humans, underlie representation and binding of spatio-temporal memory? Overall, our proposed experiments will provide novel insight into the neural basis of episodic memory as they take a new approach to this issue paradigmatically and methodologically and allow us to test several different models of hippocampal function, including our model. Neurodegenerative diseases such as stroke, epilepsy, and schizophrenia impact hippocampal subregion function and coordinated neural activity there, often resulting in symptoms of spatial disorientation and temporal confusion in patients afflicted with these conditions. Our approach thus will also have significant implications for neural diseases that affect the hippocampus.

The goal of this project is to determine the neural basis of the spatial and temporal components that comprise human episodic memory and navigation. Damage to the human hippocampus results in significant impairments to both episodic memory and navigation yet the commonalities behaviorally and neurally remain unclear. We hypothesize that spatial and temporal contextual representations, which in turn include temporal order and interval, underlie episodic memory and navigation in both partially overlapping and unique manners. To understand how the hippocampus codes spatial and temporal context, Aim 1 focuses on employing high-resolution hippocampal functional magnetic resonance imaging (fMRI) and intracranial encephalography (iEEG) to better understand the specific contributions of the microcircuitry of the human hippocampus. Building on experiments and a model we have developed in the past funding period, we hypothesize that hippocampal subfields CA3/DG play a role in differentiation of spatial vs. temporal context while CA1 plays a role in integrating commonalities across these two different forms of context. High-resolution hippocampal fMRI experiments directly test these ideas by employing a combination of experimental designs to tease apart spatial and temporal processing coupled with multivariate pattern analyses (MVPA) to map hippocampal distributed codes for these behavioral components. Hippocampal iEEG experiments focus on understanding how low- frequencies oscillations code both spatial distance and temporal contexts, particularly temporal intervals, which we hypothesize relates primarily to differences in the frequencies of oscillations. Aim 2 provides a more ?macro? perspective on human episodic memory and navigation, with a focus on the unique cortical-hippocampal and cortical-cortical networks that comprise spatial vs. temporal (order and interval) contextual processing. Building on experiments and a model we have developed over the past funding period, we will employ both whole brain fMRI and multilobular iEEG recordings in patients undergoing seizure monitoring to determine the unique cortical contributions to spatial vs. temporal context. We hypothesize that unique configurations of networks and frequencies of interactions, such as prefrontal-hippocampal interactions for temporal context and parietal-retrosplenial-hippocampal interactions for spatial context, are critical to these representations. Proposed experiments directly test these ideas by again employing both episodic memory and navigation related paradigms. The expected outcomes from this proposal are a better understanding, at both the micro and macro level scale, of how spatial vs. temporal context contribute to human episodic memory and navigation. Specifically, by better understanding the contributions of the hippocampal circuitry to episodic memory and navigation, we can better understand how diseases like stroke and ischemia impact function there. In addition, by delineating the extra-hippocampal cortical contributions, we can better understand and predict compensation following insults to the hippocampus.

Project Abstract Lesions to the human medial temporal result in striking and often long-lasting deficits in delayed verbal memory, termed medial temporal lobe amnesia. While models of medial temporal lobe function, such as declarative memory theory, hypothesize a restricted role for the medial temporal lobe in memory function, a growing consensus in cognitive neuroscience suggests that such deficits also include impairments in representations important to other areas of cognition, such as perception, attention, working memory, language, and spatial navigation. Here, we propose a novel model to better account for the range of cognitive deficits that accompany medial temporal lobe lesions. Our model hypothesizes that human medial temporal lobe function can best be described as involving both representational precision and binding, predicting increasing deficits as task-demands increase along these two critical dimensions. Experiments in Aim 1 test our model with a particular focus on testing representational precision in the context of memory and navigation to allow us to compare the outcomes from these experiments against those predicted by declarative memory theory. Experiments will include testing with bilateral medial temporal lobe patients, including those with lesions primarily restricted to the hippocampus, and high-resolution fMRI studies in healthy participants to better determine the mechanistic basis of hippocampal contributions to precision and binding. Aim 2 will determine the predictive capacity of our model, in conjunction with fMRI-based network modeling, to explain deficits accompanying unilateral medial temporal lesions that occur as a result of surgical resections during treatment of pharmacologically intractable epilepsy. The anticipated outcomes from the proposal are: 1) a more complete account of the consequences of medial temporal lobe lesions, particularly to the hippocampus, on cognition than can be provided by neuropsychological measures alone 2) a more complete predictive model of the effects of unilateral temporal lobe resection on cognitive outcomes post-resection, possibly allowing greater flexibility in determining which patients should undergo responsive neurostimulation (RNS) vs. resection 3) modeling whether and how extra-medial temporal lobe cortical networks can compensate for lost function following resection 4) potentially, inspiration for novel therapies involving cognitive interventions or neurostimulation targeting intact cortical tissue in patients with amnestic-like symptoms.

Project Abstract A cornerstone of cognitive neuroscience involves the idea that cognitive expertise can be tracked through focal changes in gray matter. One proposed mechanism for how this could work is that changes in synaptic plasticity result in dendritic growth, which in turn result in volumetric increase in gray matter observable with MRI. Consistent with this, one highly influential study suggested that taxi-drivers, who routinely employ cognitive maps and their memory of environments to navigate, have an enlarged posterior hippocampus compared to healthy control subjects and bus-drivers. A recent large- sample study from co-PI Weisberg, however, found no correlation between hippocampal volume and navigational performance in a city-like virtual reality task. Recent work has also cast doubt on whether gray matter volume changes are an appropriate measure of plasticity as they also likely involve changes in vascularization and other difficult to isolate factors. Both task-related functional and resting state connectivity offer a novel and powerful means of assaying brain wide changes potentially better related to plasticity. Such measures could also arguably be better candidates for tracking changes in skill acquisition. Here, we propose to resolve the issues above, and additionally attempt to separate navigation vs. memory functions, by having one group of participants undergo intensive training in orientation and another group undergo intensive training in episodic memory (Aim 1). We will obtain pre- and post-training measures of structural brain volume, task-related functional connectivity, and resting state connectivity to determine whether and how novel cognitive skill acquisition affects these neural measures. In addition, we will collect structural brain scans and behavioral measures from published studies to attempt understand what brain regions correlate with navigation and memory performance (Aim 2). This will allow us to perform a meta- analysis of a large sample of studies to determine how regional brain volume correlates with individual variability in these two important cognitive functions. The expected outcomes of this proposal are a better understanding of how focal gray matter vs. connectivity, as measured with MRI, relate to memory vs. navigation skills. Such outcomes could influence cognitive or stimulation therapies for stroke patients by providing insight into what brain regions or networks to target to mitigate cognitive decline.