Jean-Marc Fellous
- Professor, Psychology
- Professor, Applied Mathematics - GIDP
- Professor, Neuroscience - GIDP
- Professor, Biomedical Engineering
- Professor, Cognitive Science - GIDP
- Professor, Physiological Sciences - GIDP
- Member of the Graduate Faculty
Contact
- (520) 621-7447
- Psychology, Rm. 312
- Tucson, AZ 85721
- fellous@arizona.edu
Degrees
- Ph.D. Computer Science - Artificial Intelligence
- University of Southern California, Los Angeles, California
- A Neural Code for Face Representation: from V1 Receptive Fields to IT 'Face Cells'
Work Experience
- National Institute of Mental Health (2019 - 2020)
- Departent of Psychology and Biomedical Engineering (2017 - Ongoing)
- Department of Psychology - University of Arizona (2007 - 2017)
- Duke University (2004 - 2006)
Interests
Teaching
Computational neuroscience, Neurobiology, Data analyses
Research
Computational neuroscience, Neurobiology, Artificial Intelligence
Courses
2024-25 Courses
-
Comp Neurosci - Multi Units
PSY 535 (Fall 2024) -
Comp. Neurosci: Multi Units
PSY 435 (Fall 2024) -
Thesis
BME 910 (Fall 2024)
2023-24 Courses
-
Directed Research
PSYS 392 (Spring 2024) -
Directed Research
PSYS 492 (Spring 2024) -
Honors Directed Research
NROS 492H (Spring 2024) -
Honors Thesis
NROS 498H (Spring 2024) -
Independent Study
PSY 399 (Spring 2024) -
Systems Neuroscience
NRSC 560 (Spring 2024) -
Thesis
BME 910 (Spring 2024) -
Directed Research
PSYS 392 (Fall 2023) -
Honors Thesis
NROS 498H (Fall 2023) -
Independent Study
NROS 399 (Fall 2023) -
Intro to Biopsychology
PSY 302 (Fall 2023) -
Preceptorship
PSY 391 (Fall 2023) -
Rsrch Meth Biomed Engr
BME 592 (Fall 2023)
2022-23 Courses
-
Directed Research
PSYS 392 (Spring 2023) -
Honors Directed Research
PSYS 492H (Spring 2023) -
Honors Thesis
NSCS 498H (Spring 2023) -
Mind and Brain
PSY 300 (Spring 2023) -
Systems Neuroscience
NRSC 560 (Spring 2023) -
Thesis
BME 910 (Spring 2023) -
Honors Thesis
NSCS 498H (Fall 2022) -
Intro Computatnl Neurosc
PSY 403C (Fall 2022) -
Intro Computatnl Neurosc
PSY 503C (Fall 2022) -
Intro to Biopsychology
PSY 302 (Fall 2022) -
Preceptorship
PSY 391 (Fall 2022)
2021-22 Courses
-
Intro Neural Data Anlys
PSY 596L (Spring 2022) -
Intro to Biopsychology
PSY 302 (Spring 2022) -
Systems Neuroscience
NRSC 560 (Spring 2022)
2020-21 Courses
-
Honors Independent Study
NSCS 399H (Spring 2021) -
Honors Thesis
NSCS 498H (Spring 2021) -
Methods In Neuroscience
NRSC 700 (Spring 2021) -
Research
PS 900 (Spring 2021) -
Systems Neuroscience
NRSC 560 (Spring 2021) -
Directed Research
MCB 792 (Fall 2020) -
Honors Independent Study
MCB 399H (Fall 2020) -
Honors Independent Study
NSCS 399H (Fall 2020) -
Honors Thesis
NSCS 498H (Fall 2020) -
Independent Study
PSY 599 (Fall 2020) -
Intro Computatnl Neurosc
PSY 403C (Fall 2020) -
Intro Computatnl Neurosc
PSY 503C (Fall 2020) -
Rsrch Meth Psio Sci
PS 700 (Fall 2020)
2019-20 Courses
-
Research
PS 900 (Spring 2020) -
Systems Neuroscience
NRSC 560 (Spring 2020) -
Directed Research
PSYS 492 (Fall 2019) -
Dissertation
PSY 920 (Fall 2019) -
Honors Independent Study
BME 299H (Fall 2019) -
Research
PS 900 (Fall 2019)
2018-19 Courses
-
Directed Research
NSCS 392 (Spring 2019) -
Dissertation
PSY 920 (Spring 2019) -
Honors Thesis
NSCS 498H (Spring 2019) -
Research
PS 900 (Spring 2019) -
Systems Neuroscience
NRSC 560 (Spring 2019) -
Dissertation
PSY 920 (Fall 2018) -
Honors Independent Study
NSCS 499H (Fall 2018) -
Honors Thesis
NSCS 498H (Fall 2018) -
Intro Neural Data Anlys
PSY 496L (Fall 2018) -
Intro Neural Data Anlys
PSY 596L (Fall 2018) -
Rsrch Meth Psio Sci
PS 700 (Fall 2018)
2017-18 Courses
-
Directed Research
BME 492 (Spring 2018) -
Dissertation
PSY 920 (Spring 2018) -
Honors Independent Study
NSCS 399H (Spring 2018) -
Honors Thesis
NSCS 498H (Spring 2018) -
Honors Thesis
PSY 498H (Spring 2018) -
Independent Study
NSCS 399 (Spring 2018) -
Journal Club
APPL 595B (Spring 2018) -
Systems Neuroscience
NRSC 560 (Spring 2018) -
Directed Research
BME 492 (Fall 2017) -
Dissertation
PSY 920 (Fall 2017) -
Honors Independent Study
NSCS 399H (Fall 2017) -
Honors Thesis
NSCS 498H (Fall 2017) -
Intro to Biopsychology
PSY 302 (Fall 2017) -
Methods In Neuroscience
NRSC 700 (Fall 2017)
2016-17 Courses
-
Dissertation
PSY 920 (Spring 2017) -
Independent Study
PSY 499 (Spring 2017) -
Journal Club
APPL 595B (Spring 2017) -
Systems Neuroscience
NRSC 560 (Spring 2017) -
Thesis
PSY 910 (Spring 2017) -
Honors Independent Study
NSCS 399H (Fall 2016) -
Honors Independent Study
PSY 499H (Fall 2016) -
Independent Study
PSY 599 (Fall 2016) -
Intro to Biopsychology
PSY 302 (Fall 2016) -
Journal Club
APPL 595B (Fall 2016) -
Master's Report
PSY 909 (Fall 2016) -
Thesis
PSY 910 (Fall 2016)
2015-16 Courses
-
Directed Research
NSCS 392 (Spring 2016) -
Honors Thesis
NSCS 498H (Spring 2016) -
Journal Club
APPL 595B (Spring 2016) -
Research
PSY 900 (Spring 2016) -
Systems Neuroscience
NRSC 560 (Spring 2016)
Scholarly Contributions
Books
- Navratilova, Z., Fellous, J., & Fellous, J. M. (2007). A Biophysical Model of Cortical Up and Down States: Excitatory-Inhibitory Balance and H-Current. Springer, Berlin, Heidelberg. doi:10.1007/978-3-540-88853-6_5More infoDuring slow-wave sleep, cortical neurons oscillate between up and down states. Using a computational model of cortical neurons with realistic synaptic transmission, we determined that reverberation of activity in a small network of about 40 pyramidal cells could account for the properties of up states in vivo. We found that experimentally accessible quantities such as membrane potential fluctuations, firing rates and up state durations could be used as indicators of the size of the network undergoing the up state. We also show that the H-current, together with feed-forward inhibition can act as a gating mechanism for up state initiation.
- Fellous, J., Arbib, M. A., Arbib, M. A., & Fellous, J. M. (2005). Who Needs Emotions?: The Brain Meets the Robot. Oxford University Press. doi:10.1093/ACPROF:OSO/9780195166194.001.0001More infoPART 1: PERSPECTIVES 1. "Edison" and "Russell": Definition versus inventions in the analysis of emotion 2. Could a robot have emotions? Theoretical perspectives from social cognitive neuroscience PART 2: BRAINS 3. Neurochemical networks encoding emotion and motivation: An evolutionary perspective 4. Towards basic principles for emotional processing: What the fearful brain tells the robot 5. What are emotions, why do we have emotions, and what is their computational basis in the brain? 6. How do we decipher others' minds? PART 3: ROBOTS 7. Affect and proto-affect in effective functioning 8. The architectural basis of affective states and processes 9. Moving up the food chain: Motivation and emotion in behaviour-based robots 10. Robot emotion: A functional perspective 11. The role of emotions in multiagent teamwork PART 4: CONCLUSIONS 12. Beware the passionate robot
Chapters
- Harland, B., Contreras, M., & Fellous, J. (2018). A Role for the Longitudinal Axis of the Hippocampus in Multiscale Representations of Large and Complex Spatial Environments and Mnemonic Hierarchies. In Hippocampus: Plasticity and Functions. IntechOpen. doi:10.5772/intechopen.68877
- Fellous, J., Canavier, C., & Hasselmo, M. (2015). Neuromodulation: Increasing Computational Power. In From Brain to Cognition via Computational Neuroscience.More info(M.A. Arbib, editor)
- Fellous, J., Ledoux, J. E., & Fellous, J. M. (2012). Toward Basic Principles for Emotional Processing. In Who Needs Emotions.(pp -). -: Oxford University Press. doi:10.1093/ACPROF:OSO/9780195166194.003.0004More infoThis chapter examines the basic principles governing emotional processing. It re-evaluates the concept of the limbic system and identifies the amygdala as a crucial component of the system involved in the acquisition, storage, and expression of fear memory. Amygdala acts as a species-specific danger detector that can be quickly activated by threatening stimuli, and that can be modulated by higher cognitive systems. The amygdala influences the cognitive system by way of projections to arousal centres that control the way actions and perceptions are performed.
- Fellous, J., & Fellous, J. M. (2009). Emotion: Computational Modeling. In Encyclopedia of Neuroscience. Elsevier Ltd. doi:10.1016/B978-008045046-9.01845-3More infoThe neural basis of human emotions is difficult to study, because emotions are primarily subjective and nondeterministic. To find basic principles of emotions and their underlying mechanisms, neuroscientists typically study specific emotions, using specific tasks. They use a combination of animal and human preparations, yielding various types of data, from single neuron firing patterns, to activation levels of a whole brain area. The approach, while rigorous, is slow and yields an increasingly complex body of often conflicting data. An integrative approach is needed. As described in this article, computational models of emotion have emerged as a promising tool for integration. Because these models require that all assumptions be made explicit, they offer a new language in which to express and test hypotheses and to explain and predict neural mechanisms.
- Fellous, J., Sejnowski, T. J., Sejnowski, T. J., Navratilova, Z., & Fellous, J. M. (2009). Intrinsic and Network Contributions to Reverberatory Activity: Reactive Clamp and Modeling Studies. In Dynamic Clamp. Springer, New York, NY. doi:10.1007/978-0-387-89279-5_11More infoCortical cells belong to small interconnected ensembles. These ensembles have the potential of being activated in a reverberatory fashion in vitro and in vivo, spontaneously or in response to stimulation. We combined computer simulations and in vitro intracellular recording from prefrontal cortical neurons to explore the elicitation, modulation, and termination of these reverberations. In computer simulations, we studied the reverberating activity of small networks of neurons connected with realistic stochastic synaptic transmission and concluded that about 40 excitatory cells and a few interneurons were sufficient to reproduce the membrane and firing characteristics observed in vivo. Using a variant of the dynamic-clamp technique in vitro, we then stimulated the assembly and triggered self-sustained activity mimicking the activity recorded during the delay period of a working memory task in the behaving monkey. The onset of sustained activity depended on the number of action potentials elicited by the cue-like stimulation. Too few spikes failed to provide enough NMDA current to drive sustained reverberations; too many spikes activated a slow intrinsic hyperpolarizing current that prevented spiking; an intermediate number of spikes produced sustained activity. The firing rate during the delay period could be effectively modulated by the standard deviation of the inhibitory background synaptic noise without significant changes in the background firing rate before cue-onset. These results suggest that the balance between fast feedback inhibition and slower AMPA and NMDA feedback excitation is critical in initiating persistent activity, that intrinsic currents may determine which cell contributes to the onset or offset of reverberations and that the maintenance of persistent activity may be regulated by the amount of correlated background inhibition.
- Fellous, J., & Ledoux, J. E. (2005). Toward basic principles for emotional processing: What the fearful brain tells the robot. In Who Needs Emotions. Oxford University Press.
- Fellous, J., Arbib, M. A., Arbib, M. A., & Fellous, J. M. (2005). "Edison" and "Russell". In Who Needs Emotions. Oxford University Press. doi:10.1093/ACPROF:OSO/9780195166194.003.0001
- Fellous, J., Jonhston, T., Segal, M. R., Lisman, J., Lisman, J. E., & Fellous, J. M. (1998). Carbachol-induced rhythms in the hippocampal slice: slow (.5-2HZ), theta (4-10HZ) and gamma (80-100HZ) oscillations. In Computational Neuroscience. Plenum Press. doi:10.1007/978-1-4615-4831-7_61More infoThe hippocampus is the locus of various in vivo brain rhythms. Early work in the freely moving rat showed that three types of hippocampal oscillatory activity could be detected, depending on the behavior of the animal1. Low frequency (.5–2 Hz) irregular oscillations predominate during slow wave sleep, and are completely absent during walking. A medium frequency (5–10 Hz) rhythmical component predominates during walking behavior or REM sleep, and is absent during slow wave sleep. Finally, a fast oscillatory component (40–100 Hz) can be observed during REM sleep or walking. The neuronal circuits involved in each of these oscillations are still largely unknown, and are likely to involve the complex interplay between intrinsic cellular and synaptic hippocampal properties, and external rhythmic inputs from subcortical areas. In particular, the septal cholinergic projection to the hippocampus has been shown to significantly contribute to some of these rhythms. Here we focus on the effects of carbachol (CCH, a cholinergic agonist) on the intrinsic hippocampal circuitry. Using field recordings, we show that rhythms in these three frequency ranges may be observed in vitro, and are therefore likely to be the result of synchronized population activity. Further work in this system will allow for a detailed exploration of the neural circuitry involved and its computational role in learning and memory.
- Arbib, M. A., Arbib, M. A., & Fellous, J. (1996). A neural code for face representation: from v1 receptive fields to it 'face' cells. In Thesis. University of Southern California.
Journals/Publications
- Fellous, J., Sharpee, T., Metzner, C., Jolivet, R. B., Haas, J. S., Albada, S. J., & Nowotny, T. (2022). Editorial: Advances in Computational Neuroscience. Frontiers in Computational Neuroscience. doi:10.3389/fncom.2021.824899
- Fellous, J., Weitzenfeld, A., & Scleidorovich, P. (2022). Adapting hippocampus multi-scale place field distributions in cluttered environments optimizes spatial navigation and learning. Frontiers in Computational Neuroscience. doi:10.3389/fncom.2022.1039822More infoExtensive studies in rodents show that place cells in the hippocampus have firing patterns that are highly correlated with the animal's location in the environment and are organized in layers of increasing field sizes or scales along its dorsoventral axis. In this study, we use a spatial cognition model to show that different field sizes could be exploited to adapt the place cell representation to different environments according to their size and complexity. Specifically, we provide an in-depth analysis of how to distribute place cell fields according to the obstacles in cluttered environments to optimize learning time and path optimality during goal-oriented spatial navigation tasks. The analysis uses a reinforcement learning (RL) model that assumes that place cells allow encoding the state. While previous studies have suggested exploiting different field sizes to represent areas requiring different spatial resolutions, our work analyzes specific distributions that adapt the representation to the environment, activating larger fields in open areas and smaller fields near goals and subgoals (e.g., obstacle corners). In addition to assessing how the multi-scale representation may be exploited in spatial navigation tasks, our analysis and results suggest place cell representations that can impact the robotics field by reducing the total number of cells for path planning without compromising the quality of the paths learned.
- Lindner, B., Thomas, P. J., Fellous, J. M., & Tiesinga, P. (2021). Biological Cybernetics: 60 years and more to come. Biological cybernetics, 115(1), 5-6.
- Souder, M., Harland, B., Fellous, J. M., & Contreras, M. (2021). Dorsal CA1 hippocampal place cells form a multi-scale representation of megaspace.. BioArchives, 31(10), 2178-2190.e6. doi:10.1016/j.cub.2021.03.003More infoSpatially firing "place cells" within the hippocampal CA1 region form internal maps of the environment necessary for navigation and memory. In rodents, these neurons have been almost exclusively studied in small environments (
- Souder, M., Harland, B., Fellous, J. M., & Contreras, M. (2021). Dorsal CA1 hippocampal place cells form a multi-scale representation of megaspace.. Current biology : CB, 31(10), 2178-2190.e6. doi:10.1016/j.cub.2021.03.003More infoSpatially firing "place cells" within the hippocampal CA1 region form internal maps of the environment necessary for navigation and memory. In rodents, these neurons have been almost exclusively studied in small environments (
- Chou, Y., Rapcsak, S., Fellous, J., Wilson, R. C., Sundman, M. H., Lim, K., Ton That, V., Mizell, J., Ugonna, C., Rodriguez, R., Chen, N., Fuglevand, A. J., & Liu, Y. (2020). Transcranial magnetic stimulation reveals diminished homoeostatic metaplasticity in cognitively impaired adults. Brain Communications, 2(2). doi:10.1093/braincomms/fcaa203
- Fellous, J. M., Dominey, P., & Weitzenfeld, A. (2020). Complex spatial navigation in animals, computational models and neuro-inspired robots. Biological cybernetics, 114(2), 137-138.
- Fellous, J., Lin, K., & Xiao, Z. (2020). Conjunctive reward-place coding properties of dorsal distal CA1 hippocampus cells. Biological Cybernetics, 114, 285-301. doi:10.1007/s00422-020-00830-0
- Scleidorovich, P., Llofriu, M., Fellous, J. M., & Weitzenfeld, A. (2020). A computational model for spatial cognition combining dorsal and ventral hippocampal place field maps: multiscale navigation. Biological cybernetics.More infoClassic studies have shown that place cells are organized along the dorsoventral axis of the hippocampus according to their field size, with dorsal hippocampal place cells having smaller field sizes than ventral place cells. Studies have also suggested that dorsal place cells are primarily involved in spatial navigation, while ventral place cells are primarily involved in context and emotional encoding. Additionally, recent work has shown that the entire longitudinal axis of the hippocampus may be involved in navigation. Based on the latter, in this paper we present a spatial cognition reinforcement learning model inspired by the multiscale organization of the dorsal-ventral axis of the hippocampus. The model analyzes possible benefits of a multiscale architecture in terms of the learning speed, the path optimality, and the number of cells in the context of spatial navigation. The model is evaluated in a goal-oriented task where simulated rats need to learn a path to the goal from multiple starting locations in various open-field maze configurations. The results show that smaller scales of representation are useful for improving path optimality, whereas larger scales are useful for reducing learning time and the number of cells required. The results also show that combining scales can enhance the performance of the multiscale model, with a trade-off between path optimality and learning time.
- Sundman, M. H., Lim, K., Ton That, V., Mizell, J. M., Ugonna, C., Rodriguez, R., Chen, N. K., Fuglevand, A. J., Liu, Y., Wilson, R. C., Fellous, J. M., Rapcsak, S., & Chou, Y. H. (2020). Transcranial magnetic stimulation reveals diminished homoeostatic metaplasticity in cognitively impaired adults. Brain communications, 2(2), fcaa203.More infoHomoeostatic metaplasticity is a neuroprotective physiological feature that counterbalances Hebbian forms of plasticity to prevent network destabilization and hyperexcitability. Recent animal models highlight dysfunctional homoeostatic metaplasticity in the pathogenesis of Alzheimer's disease. However, the association between homoeostatic metaplasticity and cognitive status has not been systematically characterized in either demented or non-demented human populations, and the potential value of homoeostatic metaplasticity as an early biomarker of cognitive impairment has not been explored in humans. Here, we report that, through pre-conditioning the synaptic activity prior to non-invasive brain stimulation, the association between homoeostatic metaplasticity and cognitive status could be established in a population of non-demented human subjects (older adults across cognitive spectrums; all within the non-demented range). All participants ( = 40; age range, 65-74, 47.5% female) underwent a standardized neuropsychological battery, magnetic resonance imaging and a transcranial magnetic stimulation protocol. Specifically, we sampled motor-evoked potentials with an input/output curve immediately before and after repetitive transcranial magnetic stimulation to assess neural plasticity with two experimental paradigms: one with voluntary muscle contraction (i.e. modulated synaptic activity history) to deliberately introduce homoeostatic interference, and one without to serve as a control condition. From comparing neuroplastic responses across these experimental paradigms and across cohorts grouped by cognitive status, we found that (i) homoeostatic metaplasticity is diminished in our cohort of cognitively impaired older adults and (ii) this neuroprotective feature remains intact in cognitively normal participants. This novel finding suggests that (i) future studies should expand their scope beyond just Hebbian forms of plasticity that are traditionally assessed when using non-invasive brain stimulation to investigate cognitive ageing and (ii) the potential value of homoeostatic metaplasticity in serving as a biomarker for cognitive impairment should be further explored.
- Thomas, P. J., Fellous, J. M., & Benjamin, L. (2020). A Renewed Vision for Biological Cybernetics.. Biological cybernetics, 114(3), 315-316. doi:10.1007/s00422-020-00837-7
- Cazin, N., Llofriu Alonso, M., Scleidorovich Chiodi, P., Pelc, T., Harland, B., Weitzenfeld, A., Fellous, J. M., & Dominey, P. F. (2019). Reservoir computing model of prefrontal cortex creates novel combinations of previous navigation sequences from hippocampal place-cell replay with spatial reward propagation. PLoS computational biology, 15(7), e1006624.More infoAs rats learn to search for multiple sources of food or water in a complex environment, they generate increasingly efficient trajectories between reward sites. Such spatial navigation capacity involves the replay of hippocampal place-cells during awake states, generating small sequences of spatially related place-cell activity that we call "snippets". These snippets occur primarily during sharp-wave-ripples (SWRs). Here we focus on the role of such replay events, as the animal is learning a traveling salesperson task (TSP) across multiple trials. We hypothesize that snippet replay generates synthetic data that can substantially expand and restructure the experience available and make learning more optimal. We developed a model of snippet generation that is modulated by reward, propagated in the forward and reverse directions. This implements a form of spatial credit assignment for reinforcement learning. We use a biologically motivated computational framework known as 'reservoir computing' to model prefrontal cortex (PFC) in sequence learning, in which large pools of prewired neural elements process information dynamically through reverberations. This PFC model consolidates snippets into larger spatial sequences that may be later recalled by subsets of the original sequences. Our simulation experiments provide neurophysiological explanations for two pertinent observations related to navigation. Reward modulation allows the system to reject non-optimal segments of experienced trajectories, and reverse replay allows the system to "learn" trajectories that it has not physically experienced, both of which significantly contribute to the TSP behavior.
- Fellous, J. M., Sapiro, G., Rossi, A., Mayberg, H., & Ferrante, M. (2019). Explainable Artificial Intelligence for Neuroscience: Behavioral Neurostimulation. Frontiers in neuroscience, 13, 1346.More infoThe use of Artificial Intelligence and machine learning in basic research and clinical neuroscience is increasing. AI methods enable the interpretation of large multimodal datasets that can provide unbiased insights into the fundamental principles of brain function, potentially paving the way for earlier and more accurate detection of brain disorders and better informed intervention protocols. Despite AI's ability to create accurate predictions and classifications, in most cases it lacks the ability to provide a mechanistic understanding of how inputs and outputs relate to each other. Explainable Artificial Intelligence (XAI) is a new set of techniques that attempts to provide such an understanding, here we report on some of these practical approaches. We discuss the potential value of XAI to the field of neurostimulation for both basic scientific inquiry and therapeutic purposes, as well as, outstanding questions and obstacles to the success of the XAI approach.
- Harper, B., & Fellous, J. M. (2019). Ground truth construction and parameter tuning for the detection of sleep spindle timing in rodents. Journal of neuroscience methods, 313, 13-23.More infoThe precise detection of cortical sleep spindles is critical to basic research on memory consolidation in rodents. Previous research using automatic spindle detection algorithms often lacks systematic parameter variations and validations.
- Contreras, M., Pelc, T., Llofriu, M., Weitzenfeld, A., & Fellous, J. M. (2018). The ventral hippocampus is involved in multi-goal obstacle-rich spatial navigation. Hippocampus, 28(12), 853-866.More infoA large body of evidence shows that the hippocampus is necessary for successful spatial navigation. Various studies have shown anatomical and functional differences between the dorsal (DHC) and ventral (VHC) portions of this structure. The DHC is primarily involved in spatial navigation and contains cells with small place fields. The VHC is primarily involved in context and emotional encoding contains cells with large place fields and receives major projections from the medial prefrontal cortex. In the past, spatial navigation experiments have used relatively simple tasks that may not have required a strong coordination along the dorsoventral hippocampal axis. In this study, we tested the hypothesis that the DHC and VHC may be critical for goal-directed navigation in obstacle-rich environments. We used a learning task in which animals memorize the location of a set of rewarded feeders, and recall these locations in the presence of small or large obstacles. We report that bilateral DHC or VHC inactivation impaired spatial navigation in both large and small obstacle conditions. Importantly, this impairment did not result from a deficit in the spatial memory for the set of feeders (i.e., recognition of the goal locations) because DHC or VHC inactivation did not affect recall performance when there was no obstacle on the maze. We also show that the behavioral performance of the animals was correlated with several measures of maze complexity and that these correlations were significantly affected by inactivation only in the large object condition. These results suggest that as the complexity of the environment increases, both DHC and VHC are required for spatial navigation.
- Lindner, B., Thomas, P. J., & Fellous, J. M. (2018). Welcome from the new Editor(s)-in-Chief. Biological cybernetics, 112(3), 163.
- Fellous, J., Nation, K., & Lines, J. (2017). Dorsoventral and Proximodistal Hippocampal Processing Account for the Influences of Sleep and Context on Memory (Re)consolidation: A Connectionist Model. Computational Intelligence and Neuroscience, 2017, 1-16. doi:10.1155/2017/8091780
- Gianelli, S., Harland, B., & Fellous, J. M. (2018). A new rat-compatible robotic framework for spatial navigation behavioral experiments. Journal of neuroscience methods, 294, 40-50.More infoUnderstanding the neural substrate of information encoding and processing requires a precise control of the animal's behavior. Most of what has been learned from the rodent navigational system results from relatively simple tasks in which the movements of the animal is controlled by corridors or walkways, passive movements, treadmills or virtual reality environments. While a lot has been and continues to be learned from these types of experiments, recent evidence has shown that such artificial constraints may have significant consequences on the functioning of the neural circuits of spatial navigation.
- Lines, J., Nation, K., & Fellous, J. (2017). Dorsoventral and Proximodistal Hippocampal Processing Account for the Influences of Sleep and Context on Memory (Re) consolidation: A Connectionist Model. COMPUTATIONAL INTELLIGENCE AND NEUROSCIENCE.
- Corral-Frías, N. S., Nadel, L., Fellous, J. M., & Jacobs, W. J. (2016). Behavioral and self-reported sensitivity to reward are linked to stress-related differences in positive affect. Psychoneuroendocrinology, 66, 205-13.More infoDespite the high prevalence of stress exposure healthy adaptation or resilience is a common response. Theoretical work and recent empirical evidence suggest that a robust reward system, in part, supports healthy adaptation by preserving positive emotions even under exceptionally stressful circumstances. We tested this prediction by examining empirical relations among behavioral and self-reported measures of sensitivity to reward, trait resilience, and measures of affect in the context of experimentally induced stress. Using a quasi-experimental design we obtained measures of sensitivity to reward (self-report and behavioral), as well as affective and physiological responses to experimental psychosocial stress in a sample of 140 healthy college-age participants. We used regression-based moderation and mediational models to assess associations among sensitivity to reward, affect in the context of stress, and trait resilience and found that an interaction between exposure to experimental stress and self-reported sensitivity to reward predicted positive affect following experimental procedure. Participants with high sensitivity to reward reported higher positive affect following stress. Moreover, positive affect during or after stress mediated the relation between sensitivity to reward and trait resilience. Consistent with the prediction that a robust reward system serves as a protective factor against stress-related negative outcomes, our results found predictive associations among sensitivity to reward, positive affect, and resilience.
- Corral-Frías, N., Nadel, L., Fellous, J., & Jacobs, W. J. (2016). Behavioral and self-reported sensitivity to reward are linked to stress-related differences in positive affect.. Psychoneuroendocrinology, 16.
- Janezic, E. M., Uppalapati, S., Nagl, S., Contreras, M., French, E. D., & Fellous, J. M. (2016). Beneficial effects of chronic oxytocin administration and social co-housing in a rodent model of post-traumatic stress disorder. Behavioural pharmacology, 27(8), 704-717.More infoPost-traumatic stress disorder (PTSD) is in part due to a deficit in memory consolidation and extinction. Oxytocin (OXT) has anxiolytic effects and promotes prosocial behaviors in both rodents and humans, and evidence suggests that it plays a role in memory consolidation. We studied the effects of administered OXT and social co-housing in a rodent model of PTSD. Acute OXT yielded a short-term increase in the recall of the traumatic memory if administered immediately after trauma. Low doses of OXT delivered chronically had a cumulating anxiolytic effect that became apparent after 4 days and persisted. Repeated injections of OXT after short re-exposures to the trauma apparatus yielded a long-term reduction in anxiety. Co-housing with naive nonshocked animals decreased the memory of the traumatic context compared with single-housed animals. In the long term, these animals showed less thigmotaxis and increased interest in novel objects, and a low OXT plasma level. Co-housed PTSD animals showed an increase in risk-taking behavior. These results suggest beneficial effects of OXT if administered chronically through increases in memory consolidation after re-exposure to a safe trauma context. We also show differences between the benefits of social co-housing with naive rats and co-housing with other shocked animals on trauma-induced long-term anxiety.
- Malerba, P., Krishnan, G. P., Fellous, J. M., & Bazhenov, M. (2016). Hippocampal CA1 Ripples as Inhibitory Transients. PLoS computational biology, 12(4), e1004880.More infoMemories are stored and consolidated as a result of a dialogue between the hippocampus and cortex during sleep. Neurons active during behavior reactivate in both structures during sleep, in conjunction with characteristic brain oscillations that may form the neural substrate of memory consolidation. In the hippocampus, replay occurs within sharp wave-ripples: short bouts of high-frequency activity in area CA1 caused by excitatory activation from area CA3. In this work, we develop a computational model of ripple generation, motivated by in vivo rat data showing that ripples have a broad frequency distribution, exponential inter-arrival times and yet highly non-variable durations. Our study predicts that ripples are not persistent oscillations but result from a transient network behavior, induced by input from CA3, in which the high frequency synchronous firing of perisomatic interneurons does not depend on the time scale of synaptic inhibition. We found that noise-induced loss of synchrony among CA1 interneurons dynamically constrains individual ripple duration. Our study proposes a novel mechanism of hippocampal ripple generation consistent with a broad range of experimental data, and highlights the role of noise in regulating the duration of input-driven oscillatory spiking in an inhibitory network.
- Fellous, J., & Corral-Frias, N. (2015). Reliability and precision are optimal for non-uniform distributions of presynaptic release probability. Journal of Biomedical Science and Engineering.
- Fellous, J., Valdés, J. L., & McNaughton, B. L. (2015). Offline reactivation of experience-dependent neuronal firing patterns in the rat ventral tegmental area. Journal of Neurophysiology, 114(2), 1183-1195. doi:10.1152/jn.00758.2014
- Jones, B. J., Pest, S. M., Vargas, I. M., Glisky, E. L., & Fellous, J. (2015). Contextual reminders fail to trigger memory reconsolidation in aged humans and rats. Neurobiology of Learning and Memory, 120, 1-7.
- Llofriu, M., Tejera, G., Contreras, M., Pelc, T., Fellous, J. M., & Weitzenfeld, A. (2015). Goal-oriented robot navigation learning using a multi-scale space representation. NEURAL NETWORKS, 72, 62-74.
- Valdes, J. L., McNaughton, B. L., & Fellous, J. (2015). Offline reactivation of experience-dependent neuronal firing patterns in the rat ventral tegmental area. JOURNAL OF NEUROPHYSIOLOGY, 114(2), 1183-1195.
- Nation, K., Lines, J., & Fellous, J. M. (2014). A connectionist model of context-based memory reconsolidation in the hippocampus: the role of sleep. BMC Neuroscience, 15(1). doi:10.1186/1471-2202-15-s1-p163More infoContext-based memory reconsolidation has been studied in human and animal models [1]. In these paradigms, subjects learn two lists of items on two different days and are asked to recall the first list on day 3. Subjects who learn the two lists in the same spatial context make significantly more errors on day 3 than subjects who learn the lists in different contexts. This result suggests that contextual information may be linked to item information during memory formation or consolidation, and that this link is responsible for intrusions of items from the second list into the first list during recall when the lists were learned in identical contexts. The neural mechanisms underlying this process are unknown, but experimental studies have suggested that the hippocampus may be critical. Experimental work has shown that the dorsal and ventral portions of the hippocampus may implement qualitatively different functions in memory and spatial navigation [2], and that the proximal and distal portions of CA1 may carry information related to self-motion and sensory perception respectively [3]. We hypothesize that the dorso-ventral and proximal-distal anatomical differentiations of this structure may explain some of the experimental data on memory reconsolidation. To test this hypothesis, we built a connectionist model of the hippocampus (figure (figure1).1). The model is implemented using EMERGENT [4]. In this model, the dorsal stream carries predominantly item information, while the ventral stream carries spatial contextual information. In both streams, the distal CA1 encodes items using inputs from the lateral entorhinal cortex, while the proximal CA1 encodes spatial context using medial entorhinal cortical inputs. We train and test the model as in the experiments. Figure 1 Model architecture. Separate and interacting ‘object’ and ‘context’ streams shown in blue and red respectively, with explicit output ‘guesses’. We found that object representation overlap as well as additional, extraneous learning can explain how context affects recall performance and produce intrusions as observed experimentally. We then selectively lesion the network to investigate which component of the hippocampus affects context based memory recall. In addition, we use the model to understand memory reactivation during sleep. Sleep is simulated by presenting small amounts of noise in the input layers. We found that this noise partially re-activated the memory representations of objects that were previously learned. These partial memories were then set as inputs and were re-processed by the network. This in turn made these memories resilient to interference from new items learned at a later time as was shown experimentally [5]. The model will allow for an investigation of why certain items have preferential memory reactivation during sleep. Furthermore, the model may be used to explain recent experimental data showing that presenting specific external stimuli during sleep may influence the memory consolidation process [6].
- Nie, Y., Fellous, J., & Tatsuno, M. (2014). Influence of External Inputs and Asymmetry of Connections on Information-Geometric Measures Involving Up to Ten Neuronal Interactions. NEURAL COMPUTATION, 26(10), 2247-2293.More infoThe investigation of neural interactions is crucial for understanding information processing in the brain. Recently an analysis method based on information geometry (IG) has gained increased attention, and the property of the pairwise IG measure has been studied extensively in relation to the two-neuron interaction. However, little is known about the property of IG measures involving more neuronal interactions. In this study, we systematically investigated the influence of external inputs and the asymmetry of connections on the IG measures in cases ranging from 1-neuron to 10-neuron interactions. First, the analytical relationship between the IG measures and external inputs was derived for a network of 10 neurons with uniform connections. Our results confirmed that the single and pairwise IG measures were good estimators of the mean background input and of the sum of the connection weights, respectively. For the IG measures involving 3 to 10 neuronal interactions, we found that the influence of external inputs was highly nonlinear. Second, by computer simulation, we extended our analytical results to asymmetric connections. For a network of 10 neurons, the simulation showed that the behavior of the IG measures in relation to external inputs was similar to the analytical solution obtained for a uniformly connected network. When the network size was increased to 1000 neurons, the influence of external inputs almost disappeared. This result suggests that all IG measures from 1-neuron to 10-neuron interactions are robust against the influence of external inputs. In addition, we investigated how the strength of asymmetry influenced the IG measures. Computer simulation of a 1000-neuron network showed that all the IG measures were robust against the modulation of the asymmetry of connections. Our results provide further support for an information-geometric approach and will provide useful insights when these IG measures are applied to real experimental spike data.
- Nie, Y., Fellous, J., & Tatsuno, M. (2014). Information-geometric measures estimate neural interactions during oscillatory brain states. FRONTIERS IN NEURAL CIRCUITS, 8.More infoThe characterization of functional network structures among multiple neurons is essential to understanding neural information processing. Information geometry (IG), a theory developed for investigating a space of probability distributions has recently been applied to spike-train analysis and has provided robust estimations of neural interactions. Although neural firing in the equilibrium state is often assumed in these studies, in reality, neural activity is non-stationary. The brain exhibits various oscillations depending on cognitive demands or when an animal is asleep. Therefore, the investigation of the IG measures during oscillatory network states is important for testing how the IG method can be applied to real neural data. Using model networks of binary neurons or more realistic spiking neurons, we studied how the single- and pairwise-IG measures were influenced by oscillatory neural activity. Two general oscillatory mechanisms, externally driven oscillations and internally induced oscillations, were considered. In both mechanisms, we found that the single-IG measure was linearly related to the magnitude of the external input, and that the pairwise-IG measure was linearly related to the sum of connection strengths between two neurons. We also observed that the pairwise-IG measure was not dependent on the oscillation frequency. These results are consistent with the previous findings that were obtained under the equilibrium conditions. Therefore, we demonstrate that the IG method provides useful insights into neural interactions under the oscillatory condition that can often be observed in the real brain.
- Song, M. R., & Fellous, J. (2014). Value Learning and Arousal in the Extinction of Probabilistic Rewards: The Role of Dopamine in a Modified Temporal Difference Model. PLOS ONE, 9(2).More infoBecause most rewarding events are probabilistic and changing, the extinction of probabilistic rewards is important for survival. It has been proposed that the extinction of probabilistic rewards depends on arousal and the amount of learning of reward values. Midbrain dopamine neurons were suggested to play a role in both arousal and learning reward values. Despite extensive research on modeling dopaminergic activity in reward learning (e.g. temporal difference models), few studies have been done on modeling its role in arousal. Although temporal difference models capture key characteristics of dopaminergic activity during the extinction of deterministic rewards, they have been less successful at simulating the extinction of probabilistic rewards. By adding an arousal signal to a temporal difference model, we were able to simulate the extinction of probabilistic rewards and its dependence on the amount of learning. Our simulations propose that arousal allows the probability of reward to have lasting effects on the updating of reward value, which slows the extinction of low probability rewards. Using this model, we predicted that, by signaling the prediction error, dopamine determines the learned reward value that has to be extinguished during extinction and participates in regulating the size of the arousal signal that controls the learning rate. These predictions were supported by pharmacological experiments in rats.
- Fellous, J., Corral-Frias, N. S., Lahood, R. P., Edelman-Vogelsang, K. E., French, E. D., & Fellous, J. -. (2013). Involvement of the ventral tegmental area in a rodent model of post-traumatic stress disorder. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology, 38(2).More infoPost-traumatic stress disorder (PTSD) is an anxiety disorder of considerable prevalence in individuals who have experienced a traumatic event. Studies of the neural substrate of this disorder have focused on the role of areas such as the hippocampus, the amygdala and the medial prefrontal cortex. We show that the ventral tegmental area (VTA), which directly modulates these areas, is part of this circuitry. Using a rat model of PTSD, we show that a brief but intense foot shock followed by three brief reminders can cause long-term behavioral changes as shown by anxiety-like, nociception, and touch-sensitivity tests. We show that an intraperitoneal injection of a dopamine (DA) antagonist or a bilateral inactivation of the VTA administered immediately before the traumatic event decrease the occurrence or intensity of these behavioral changes. Furthermore, we show that there is a significant decrease of baseline VTA dopaminergic but not GABAergic cell firing rates 2 weeks after trauma. Our data suggest that VTA DA neurons undergo long-term physiological changes after trauma and that this brain area is a crucial part of the circuits involved in PTSD symptomatology.
- Fellous, J., Greene, P., Howard, M., & Bhattacharyya, R. (2013). Hippocampal Anatomy Supports the Use of Context in Object Recognition: A Computational Model. Computational Intelligence and Neuroscience, 2013, 1-19. doi:10.1155/2013/294878
- Franklin, R. G., Zebrowitz, L. A., Fellous, J., & Lee, A. (2013). Generalizing from human facial sexual dimorphism to sex-differentiate macaques: Accuracy and cultural variation. JOURNAL OF EXPERIMENTAL SOCIAL PSYCHOLOGY, 49(3), 344-348.
- Greene, P., Howard, M., Bhattacharyya, R., & Fellous, J. (2013). Hippocampal anatomy supports the use of context in object recognition: A computational model. Computational Intelligence and Neuroscience, 2013.More infoPMID: 23781237;PMCID: PMC3677630;Abstract: The human hippocampus receives distinct signals via the lateral entorhinal cortex, typically associated with object features, and the medial entorhinal cortex, associated with spatial or contextual information. The existence of these distinct types of information calls for some means by which they can be managed in an appropriate way, by integrating them or keeping them separate as required to improve recognition. We hypothesize that several anatomical features of the hippocampus, including differentiation in connectivity between the superior/inferior blades of DG and the distal/proximal regions of CA3 and CA1, work together to play this information managing role. We construct a set of neural network models with these features and compare their recognition performance when given noisy or partial versions of contexts and their associated objects. We found that the anterior and posterior regions of the hippocampus naturally require different ratios of object and context input for optimal performance, due to the greater number of objects versus contexts. Additionally, we found that having separate processing regions in DG significantly aided recognition in situations where object inputs were degraded. However, split processing in both DG and CA3 resulted in performance tradeoffs, though the actual hippocampus may have ways of mitigating such losses. © 2013 Patrick Greene et al.
- Lyttle, D., Lyttle, D., Gereke, B., Gereke, B., Lin, K. K., Lin, K. K., Fellous, J., & Fellous, J. (2013). Spatial scale and place field stability in a grid-to-place cell model of the dorsoventral axis of the hippocampus. Hippocampus, 23(8), 729-744.More infoPMID: 23576417;Abstract: The rodent hippocampus and entorhinal cortex contain spatially modulated cells that serve as the basis for spatial coding. Both medial entorhinal grid cells and hippocampal place cells have been shown to encode spatial information across multiple spatial scales that increase along the dorsoventral axis of these structures. Place cells near the dorsal pole possess small, stable, and spatially selective firing fields, while ventral cells have larger, less stable, and less spatially selective firing fields. One possible explanation for these dorsoventral changes in place field properties is that they arise as a result of similar dorsoventral differences in the properties of the grid cell inputs to place cells. Here, we test the alternative hypothesis that dorsoventral place field differences are due to higher amounts of nonspatial inputs to ventral hippocampal cells. We use a computational model of the entorhinal-hippocampal network to assess the relative contributions of grid scale and nonspatial inputs in determining place field size and stability. In addition, we assess the consequences of grid node firing rate heterogeneity on place field stability. Our results suggest that dorsoventral differences in place cell properties can be better explained by changes in the amount of nonspatial inputs, rather than by changes in the scale of grid cell inputs, and that grid node heterogeneity may have important functional consequences. The observed gradient in field size may reflect a shift from processing primarily spatial information in the dorsal hippocampus to processing more nonspatial, contextual, and emotional information near the ventral hippocampus. © 2013 Wiley Periodicals, Inc.
- Stidd, D. A., Vogelsang, K., Krahl, S. E., Langevin, J., & Fellous, J. (2013). Amygdala Deep Brain Stimulation Is Superior to Paroxetine Treatment in a Rat Model of Posttraumatic Stress Disorder. BRAIN STIMULATION, 6(6), 837-844.
- Tejera, G., Barrera, A., Fellous, J., Llofriu, M., & Weitzenfeld, A. (2013). Spatial cognition: Robot target localization in open arenas based on rat studies. Proceedings of SPIE - The International Society for Optical Engineering, 8756.More infoAbstract: We describe our latest work in understanding spatial localization in open arenas based on rat studies and corresponding modeling with simulated and physical robots. The studies and experiments focus on goal-oriented navigation where both rats and robots exploit distal cues to localize and find a goal in an open environment. The task involves training of both rats and robots to find the shortest path to the goal from multiple starting points in the environment. The spatial cognition model is based on the rat's brain neurophysiology of the hippocampus extending previous work by analyzing granularity of localization in relation to a varying number and position of landmarks. The robot integrates internal and external information to create a topological map of the environment and to generate shortest routes to the goal through path integration. One of the critical challenges for the robot is to analyze the similarity of positions and distinguish among different locations using visual cues and previous paths followed to reach the current position. We describe the robotics architecture used to develop, simulate and experiment with physical robots. © 2013 SPIE.
- Weitzenfeld, A., Lyttle, D., Lin, K. K., & Fellous, J. M. (2013). The influence of multiple firing events on the formation and stability of activity patterns in continuous attractor networks.. BMC Neuroscience, 14(1). doi:10.1186/1471-2202-14-s1-p241More infoContinuous attractor networks have been proposed to explain a variety of phenomena, including working memory and rodent entorhinal grid cells [1,2]. Typically, such networks consist of spatially-structured lattices of neurons in one or two dimensions with long-range inhibition and short-range excitation, which causes the network activity to spontaneously self-organize into one or more "bumps" of persistent activity. In two-dimensions, the activity patterns often take the form of triangular lattices, which bear a striking resemblance to the spatial patterns observed in recordings of entorhinal grid cells [3]. Rate-based neural field models of attractor networks predict the emergence of stable, stationary patterns [2,4], whereas simulations of integrate and fire neurons suggest a more complicated dynamical picture. In particular, synaptic timescales appear to play a crucial role [5,6] in that fast synapses tend to lead to transient, local synchronous activity, thus destabilizing activity patterns. Recent work on a model of V1 suggests that such "multiple firing events" (MFEs) may be a generic, emergent feature of spiking networks [7]. In this work, we investigate whether and how MFEs may affect pattern stability in continuous attractor networks. While the stability of spiking neuron-based attractor networks have been studied previously [5,6], the models studied to date contain neither seperate excitatory and inhibitory populations nor a mixture of synaptic timescales (as would be expected in more realistic settings), and the effects of these features on pattern stability are not known. To investigate these questions, we have implemented continuous attractor networks in both one and two dimensions using spiking, integrate-and-fire neurons with conductance-based synapses and seperate excitatory and inhibitory populations. Through simulations, we systematically assess the effect of MFEs on the formation and stability of spatiotemporal activity patterns ("bumps" and "grids"). We also investigate the effects of multiple synaptic timescales, noise, and network heterogeneity on the stability of these patterns. An overarching goal of this work is to obtain insights into how real neural systems might maintain stable persistent activity states, such as those needed for accurately integrating sensorimotor information. Furthermore, as 2-D attractor networks have been incorporated into a number of recent grid cell models [2,8], and the stability of grid cell activity may have a significant effect on place fields [9], understanding the stability of the grid patterns formed by these networks is highly relevant to studies of the rat spatial navigation system.
- Fellous, J., Jones, B., Bukoski, E., Nadel, L., & Fellous, J. -. (2012). Remaking memories: reconsolidation updates positively motivated spatial memory in rats. Learning & memory (Cold Spring Harbor, N.Y.), 19(3).More infoThere is strong evidence that reactivation of a memory returns it to a labile state, initiating a restabilization process termed reconsolidation, which allows for updating of the memory. In this study we investigated reactivation-dependent updating using a new positively motivated spatial task in rodents that was designed specifically to model a human list-learning paradigm. On Day 1, rats were trained to run to three feeders (List 1) for rewards. On Day 2, rats were trained to run to three different feeders (List 2) in either the same (Reminder condition) or a different (No Reminder condition) experimental context than on Day 1. On Day 3, rats were cued to recall List 1. Rats in the Reminder condition made significantly more visits to List 2 feeders (intrusions) during List 1 recall than rats in the No Reminder condition, indicating that the reminder triggered reactivation and allowed integration of List 2 items into List 1. This reminder effect was selective for the reactivated List 1 memory, as no intrusions occurred when List 2 was recalled on Day 3. No intrusions occurred when retrieval took place in a different context from the one used at encoding, indicating that the expression of the updated memory is dependent upon the retrieval context. Finally, the level of intrusions was highest when retrieval took place immediately after List 2 learning, and generally declined when retrieval occurred 1-4 h later, indicating that the List 2 memory competed with short-term retrieval of List 1. These results demonstrate the dynamic nature of memory over time and the impact of environmental context at different stages of memory processing.
- Lyttle, D., Lin, K. K., & Fellous, J. M. (2012). Coding, stability, and non-spatial inputs in a modular grid-to-place cell model.. BMC Neuroscience, 13(1), -. doi:10.1186/1471-2202-13-s1-p141More infoGrid cells in the medial entorhinal cortex (mEC) and place cells in the hippocampus are paradigms for population coding of spatial information [1]. Both the spatiallyperiodic firing fields of grid cells and the spatially localized firing fields of place cells show systematic increases in spatial scale along the dorso-ventral axes of the mEC and hippocampus, respectively [2,3], which has led to speculation that place field size is determined simply by the spatial scale of a place cell’s grid cell inputs. However, in addition to receiving spatially-modulated entorhinal inputs, place cells receive contextual, non-spatial inputs from sources such as the amygdala and hypothalamus [4], which may be important in determining place cell firing properties. These non-spatial inputs are particularly prominent toward the ventral pole of the hippocampus [4], and thus could also play a role in producing dorso-ventral place cell differences. In order to understand the relative contributions of grid cells and non-spatial inputs in determining place field size and stability, we propose a computational model of the hippocampal-entorhinal network that includes a modular organization of grid cell inputs arranged in order of increasing spatial scale, as is seen experimentally in the mEC. Our underlying place cell model is a firing-rate based model inspired by previous work [5], in which place fields are formed via competition between place cells. We also introduce a dorsoventral gradient in the amount of non-spatial input to place cells, with ventral cells receiving more input from non-spatial sources. Finally, we introduce heterogeneity into the firing rates of grid vertices within individual grid fields. This heterogeneity is observed in experimental recordings [6] but has received relatively little attention in experimental or theoretical work, despite its potential impact on place field stability. Our main findings suggest that: 1.) For a wide range of parameters, the relative amounts of spatial and non-spatial inputs to place cells plays a more important role in determining place field size and stability than the spatial scale of grid cell inputs. This implies that the dorso-ventral gradient in place field size may reflect a dorso-ventral gradient in non-spatial inputs, rather than grid scale, and is agreement with prior suggestions of a functional distinction between the dorsal and ventral regions of the hippocampus [7]. 2.) In our model, place fields are sensitive to changes in the firing rates of the grid vertices of individual grid cells, emphasizing the potential implications of this grid field heterogeneity for place field formation and stability.
- Navratilova, Z., Giocomo, L. M., Fellous, J., Hasselmo, M. E., & McNaughton, B. L. (2012). Phase precession and variable spatial scaling in a periodic attractor map model of medial entorhinal grid cells with realistic after-spike dynamics. HIPPOCAMPUS, 22(4), 772-789.
- Navratilova, Z., Navratilova, Z., Giocomo, L. M., Giocomo, L. M., Fellous, J., Fellous, J., Hasselmo, M. E., Hasselmo, M. E., McNaughton, B. L., & McNaughton, B. L. (2012). Phase precession and variable spatial scaling in a periodic attractor map model of medial entorhinal grid cells with realistic after-spike dynamics. Hippocampus, 22(4), 772-789.More infoPMID: 21484936;Abstract: We present a model that describes the generation of the spatial (grid fields) and temporal (phase precession) properties of medial entorhinal cortical (MEC) neurons by combining network and intrinsic cellular properties. The model incorporates network architecture derived from earlier attractor map models, and is implemented in 1D for simplicity. Periodic driving of conjunctive (position × head-direction) layer-III MEC cells at theta frequency with intensity proportional to the rat's speed, moves an 'activity bump' forward in network space at a corresponding speed. The addition of prolonged excitatory currents and simple after-spike dynamics resembling those observed in MEC stellate cells (for which new data are presented) accounts for both phase precession and the change in scale of grid fields along the dorso-ventral axis of MEC. Phase precession in the model depends on both synaptic connectivity and intrinsic currents, each of which drive neural spiking either during entry into, or during exit out of a grid field. Thus, the model predicts that the slope of phase precession changes between entry into and exit out of the field. The model also exhibits independent variation in grid spatial period and grid field size, which suggests possible experimental tests of the model. © 2011 Wiley Periodicals, Inc.
- Toups, J. V., Fellous, J., Thomas, P. J., Sejnowski, T. J., & Tiesinga, P. H. (2012). Multiple Spike Time Patterns Occur at Bifurcation Points of Membrane Potential Dynamics. PLoS Computational Biology, 8(10).More infoPMID: 23093916;PMCID: PMC3475656;Abstract: The response of a neuron to repeated somatic fluctuating current injections in vitro can elicit a reliable and precisely timed sequence of action potentials. The set of responses obtained across trials can also be interpreted as the response of an ensemble of similar neurons receiving the same input, with the precise spike times representing synchronous volleys that would be effective in driving postsynaptic neurons. To study the reproducibility of the output spike times for different conditions that might occur in vivo, we somatically injected aperiodic current waveforms into cortical neurons in vitro and systematically varied the amplitude and DC offset of the fluctuations. As the amplitude of the fluctuations was increased, reliability increased and the spike times remained stable over a wide range of values. However, at specific values called bifurcation points, large shifts in the spike times were obtained in response to small changes in the stimulus, resulting in multiple spike patterns that were revealed using an unsupervised classification method. Increasing the DC offset, which mimicked an overall increase in network background activity, also revealed bifurcation points and increased the reliability. Furthermore, the spike times shifted earlier with increasing offset. Although the reliability was reduced at bifurcation points, a theoretical analysis showed that the information about the stimulus time course was increased because each of the spike time patterns contained different information about the input. © 2012 Toups et al.
- Weitzenfeld, A., Fellous, J., Barrera, A., & Tejera, G. (2012). Allothetic and idiothetic sensor fusion in rat-inspired robot localization. Proceedings of SPIE - The International Society for Optical Engineering, 8407.More infoAbstract: We describe a spatial cognition model based on the rat's brain neurophysiology as a basis for new robotic navigation architectures. The model integrates allothetic (external visual landmarks) and idiothetic (internal kinesthetic information) cues to train either rat or robot to learn a path enabling it to reach a goal from multiple starting positions. It stands in contrast to most robotic architectures based on SLAM, where a map of the environment is built to provide probabilistic localization information computed from robot odometry and landmark perception. Allothetic cues suffer in general from perceptual ambiguity when trying to distinguish between places with equivalent visual patterns, while idiothetic cues suffer from imprecise motions and limited memory recalls. We experiment with both types of cues in different maze configurations by training rats and robots to find the goal starting from a fixed location, and then testing them to reach the same target from new starting locations. We show that the robot, after having pre-explored a maze, can find a goal with improved efficiency, and is able to (1) learn the correct route to reach the goal, (2) recognize places already visited, and (3) exploit allothetic and idiothetic cues to improve on its performance. We finally contrast our biologically-inspired approach to more traditional robotic approaches and discuss current work in progress. © 2012 SPIE.
- Fellous, J., Lyttle, D., & Fellous, J. -. (2011). A new similarity measure for spike trains: sensitivity to bursts and periods of inhibition. Journal of neuroscience methods, 199(2).More infoAn important problem in neuroscience is that of constructing quantitative measures of the similarity between neural spike trains. These measures can be used, for example, to assess the reliability of the response of a single neuron to repeated stimulus presentations, or to uncover relationships in the firing patterns of multiple neurons in a population. While several similarity measures have been proposed, the extent to which they take into account various biologically important spike train features such as bursts of spikes, or periods of inactivity remains poorly understood. Here we compare these measures using tests specifically designed to assess the sensitivity to bursts and silent periods. In addition, we propose two new measures. The first is designed to detect periods of shared silence between spike trains, while the second is designed to emphasize the presence of common bursts. To assist researchers in determining which measure is best suited to their particular data analysis needs, we also show how these measures can be combined and how their parameters can be determined on the basis of physiologically relevant quantities.
- Fellous, J., de Jong, L. W., Gereke, B., Martin, G. M., & Fellous, J. -. (2011). The traveling salesrat: insights into the dynamics of efficient spatial navigation in the rodent. Journal of neural engineering, 8(6).More infoRodent spatial navigation requires the dynamic evaluation of multiple sources of information, including visual cues, self-motion signals and reward signals. The nature of the evaluation, its dynamics and the relative weighting of the multiple information streams are largely unknown and have generated many hypotheses in the field of robotics. We use the framework of the traveling salesperson problem (TSP) to study how this evaluation may be achieved. The TSP is a classical artificial intelligence NP-hard problem that requires an agent to visit a fixed set of locations once, minimizing the total distance traveled. We show that after a few trials, rats converge on a short route between rewarded food cups. We propose that this route emerges from a series of local decisions that are derived from weighing information embedded in the context of the task. We study the relative weighting of spatial and reward information and establish that, in the conditions of this experiment, when the contingencies are not in conflict, rats choose the spatial or reward optimal solution. There was a trend toward a preference for space when the contingencies were in conflict. We also show that the spatial decision about which cup to go to next is biased by the orientation of the animal. Reward contingencies are also shown to significantly and dynamically modulate the decision-making process. This paradigm will allow for further neurophysiological studies aimed at understanding the synergistic role of brain areas involved in planning, reward processing and spatial navigation. These insights will in turn suggest new neural-like architectures for the control of mobile autonomous robots.
- Howard, M. D., O'Reilly, R. C., Ascoli, G., & Fellous, J. (2011). Adaptive recall in hippocampus. Frontiers in Artificial Intelligence and Applications, 233, 151-157.More infoAbstract: Complementary learning systems (CLS) theory describes how the hippocampal and cortical contributions to recognition memory are a direct result of their architectural and computational specializations. In this paper we model a further refinement of CLS that features separate handling of inputs from the dorsal and ventral posterior cortices, and present a possible mechanism for adaptive recall in hippocampus based on several research findings that have not previously been related to each other. This model suggests how we are able to recognize familiar objects in unfamiliar settings. © 2011 The authors and IOS Press. All rights reserved.
- Lyttle, D., & Fellous, J. (2011). A new similarity measure for spike trains: Sensitivity to bursts and periods of inhibition. Journal of Neuroscience Methods, 199(2), 296-309.More infoPMID: 21600921;Abstract: An important problem in neuroscience is that of constructing quantitative measures of the similarity between neural spike trains. These measures can be used, for example, to assess the reliability of the response of a single neuron to repeated stimulus presentations, or to uncover relationships in the firing patterns of multiple neurons in a population. While several similarity measures have been proposed, the extent to which they take into account various biologically important spike train features such as bursts of spikes, or periods of inactivity remains poorly understood. Here we compare these measures using tests specifically designed to assess the sensitivity to bursts and silent periods. In addition, we propose two new measures. The first is designed to detect periods of shared silence between spike trains, while the second is designed to emphasize the presence of common bursts. To assist researchers in determining which measure is best suited to their particular data analysis needs, we also show how these measures can be combined and how their parameters can be determined on the basis of physiologically relevant quantities. © 2011 Elsevier B.V.
- Toups, J. V., Fellous, J., Thomas, P. J., Sejnowski, T. J., & Tiesinga, P. H. (2011). Finding the event structure of neuronal spike trains. Neural Computation, 23(9), 2169-2208.More infoPMID: 21671786;PMCID: PMC3220920;Abstract: Neurons in sensory systems convey information about physical stimuli in their spike trains. In vitro, single neurons respond precisely and reliably to the repeated injection of the same fluctuating current, producing regions of elevated firing rate, termed events. Analysis of these spike trains reveals that multiple distinct spike patterns can be identified as trial-to-trial correlations between spike times (Fellous, Tiesinga, Thomas, & Sejnowski, 2004). Finding events in data with realistic spiking statistics is challenging because events belonging to different spike patterns may overlap. We propose a method for finding spiking events that uses contextual information to disambiguate which pattern a trial belongs to. The procedure can be applied to spike trains of the same neuron across multiple trials to detect and separate responses obtained during different brain states. The procedure can also be applied to spike trains from multiple simultaneously recorded neurons in order to identify volleys of near-synchronous activity or to distinguish between excitatory and inhibitory neurons. The procedure was tested using artificial data as well as recordings in vitro in response to fluctuating current waveforms. © 2011 Massachusetts Institute of Technology.
- Valdes, J. L., Mcnaughton, B. L., & Fellous, J. M. (2011). Experience-dependent reactivations of ventral tegmental area neurons in the rat. BMC Neuroscience, 12(1). doi:10.1186/1471-2202-12-s1-p107More infoThe hippocampus stores information during the acquisition of new memory episodes. These memories are replayed during sleep as part of a memory consolidation process. The neural mechanisms underlying these reactivations are currently under investigation. One hypothesis is that reactivation occurs as a result of local attractor dynamics within the structure in which they occur. Another possibility is that reactivation in these various areas is at least in part inherited from one or several other structures that project to them. Theoretical and experimental work on reinforcement learning have proposed many ways in which learning can be modulated by the value associated with a stimulus. Beyond initial memory acquisition, however, it is still unclear why specific memory items are consolidated and others are not [1,2]. One possibility is that, as with memory acquisition, the consolidation process is modulated by the value associated with a specific memory item. Research has shown that this value may be at least in part encoded by subcortical structures such as the ventral tegmental area (VTA) [3,4]. We provide new evidence in the rodent that 45% VTA neurons are sensitive to and selective for different types of stimuli. In three different tasks involving various amounts spatial and reward components, we show that putative dopaminergic VTA neurons strongly reactivate during a rest period following the tasks. This reactivation takes the form of population-wide activity patterns lasting from a few 100 ms up to a few seconds. In the non-spatial task, a statistical analysis of this reactivation using the explained variance measure showed that most of the reactivation relies on the activity of stimulus-sensitive neurons. Stimulus insensitive neurons exhibited significant reactivation if the task used involved motor and spatial components. VTA has widespread connections to the rest of the brain, including hippocampus and neocortex. Almost all experiments that have shown memory trace reactivation were based on tasks that involved rewards. The selective reactivation of non-GABAergic cells in the VTA suggests that reactivations in hippocampus and cortex could be modulated by the VTA. The finding that reward-sensitive neurons primarily reactivated when rewards were present during the task suggests a mechanism by which hippocampal and neocortical reactivations can be modulated by the value of the memory items they encode. In turn, cortex and hippocampus could modulate VTA activity as part of a loop in which memory content and memory value are dynamically and selectively established. Future conceptual and computational models of reinforcement learning and memory consolidation could account for the type of selectivity and dynamics of VTA neural firing we described here.
- Watkins, L., Gereke, B., Martin, G. M., & Fellous, J. (2011). The traveling salesrat: Insights into the dynamics of efficient spatial navigation in the rodent. Journal of Neural Engineering, 8(6).More infoPMID: 22056477;Abstract: Rodent spatial navigation requires the dynamic evaluation of multiple sources of information, including visual cues, self-motion signals and reward signals. The nature of the evaluation, its dynamics and the relative weighting of the multiple information streams are largely unknown and have generated many hypotheses in the field of robotics. We use the framework of the traveling salesperson problem (TSP) to study how this evaluation may be achieved. The TSP is a classical artificial intelligence NP-hard problem that requires an agent to visit a fixed set of locations once, minimizing the total distance traveled. We show that after a few trials, rats converge on a short route between rewarded food cups. We propose that this route emerges from a series of local decisions that are derived from weighing information embedded in the context of the task. We study the relative weighting of spatial and reward information and establish that, in the conditions of this experiment, when the contingencies are not in conflict, rats choose the spatial or reward optimal solution. There was a trend toward a preference for space when the contingencies were in conflict. We also show that the spatial decision about which cup to go to next is biased by the orientation of the animal. Reward contingencies are also shown to significantly and dynamically modulate the decision-making process. This paradigm will allow for further neurophysiological studies aimed at understanding the synergistic role of brain areas involved in planning, reward processing and spatial navigation. These insights will in turn suggest new neural-like architectures for the control of mobile autonomous robots. © 2011 IOP Publishing Ltd.
- Fellous, J. -. (2010). Emotion: Computational modeling. Encyclopedia of Neuroscience, 909-913.More infoAbstract: The neural basis of human emotions is difficult to study, because emotions are primarily subjective and nondeterministic. To find basic principles of emotions and their underlying mechanisms, neuroscientists typically study specific emotions, using specific tasks. They use a combination of animal and human preparations, yielding various types of data, from single neuron firing patterns, to activation levels of a whole brain area. The approach, while rigorous, is slow and yields an increasingly complex body of often conflicting data. An integrative approach is needed. As described in this article, computational models of emotion have emerged as a promising tool for integration. Because these models require that all assumptions be made explicit, they offer a new language in which to express and test hypotheses and to explain and predict neural mechanisms. © 2009 Elsevier Ltd All rights reserved.
- Lyttle, D., & Fellous, J. M. (2010). Analyzing spike train similarity measures: the effects of bursts and silence. BMC Neuroscience, 11(1). doi:10.1186/1471-2202-11-s1-p121More infoA fundamental issue in neuroscience research is that of quantifying the similarity or dissimilarity of patterns of neuronal activity, or ``spike trains.'' Several approaches to solving this problem have been proposed, in particular, a variety of spike train similarity measures have been constructed [1-7]. Some of these quantitative measures of similarity are metrics in the strict mathematical sense of the word, and all of them can be thought of as attempts to define the intuitive notion of a ``distance'' between two spike trains. In constructing or choosing a similarity measure, one faces the question of what exactly it means for two trains to be considered similar (close) or dissimilar (far apart), and how this definition of similarity is incorporated into the measure. That choice may depend explicitly upon the nature of the neural system being investigating. In any case, one is faced with the problem of deciding exactly which features of a spike train are physiologically relevant and important for encoding information. Features of interest include neuronal firing rate, spike timing, bursts, or periods of inactivity that are correlated between neurons. The existence of a wide range of possible mechanisms through which spike trains could be encoding information calls into question the idea that a single similarity measure is appropriate for all data sets or experimental conditions. It may be that the best choice of a similarity measure depends intimately upon the specific features of the spike data under analysis. To date, the extent to which different measures respond differently to various specific features of spike trains has for the most part not been explored, although a recent study [8] has compared a handful of these measures on the basis of their effectiveness in discriminating spike trains on the basis of firing rate, instantaneous firing rate and spike synchrony. In contrast, in the present study we propose a novel set of criteria along which to evaluate spike train similarity measures. In particular, we examine the sensitivity of the measures to periods of common silence and the presence and timing of bursts through a set of simple computational tests. We find that, of the measures we examined, only a few were sensitive to bursts, and only one displayed sensitivity to shared silence. None were sensitive to both bursts and silence. In light of this, we introduce a new measure designed specifically to detect and emphasize shared periods of silent inactivity, and evaluate this measure along the same criteria as the others. We find that this new measure, when combined in a natural way with a measure proposed by Schreiber et al., [5] is unique in its sensitivity to both bursts and shared periods of inactivity. Further work will involve subjecting this new measure to the sort of analysis conducted by others [8] to further elucidate its properties. Another promising avenue is to compare this new measure to others on the basis of clustering performance, as has been done with other measures [6].
- Samson, R. D., Frank, M. J., & Fellous, J. (2010). Computational models of reinforcement learning: The role of dopamine as a reward signal. Cognitive Neurodynamics, 4(2), 91-105.More infoPMID: 21629583;PMCID: PMC2866366;Abstract: Reinforcement learning is ubiquitous. Unlike other forms of learning, it involves the processing of fast yet content-poor feedback information to correct assumptions about the nature of a task or of a set of stimuli. This feedback information is often delivered as generic rewards or punishments, and has little to do with the stimulus features to be learned. How can such low-content feedback lead to such an efficient learning paradigm? Through a review of existing neuro-computational models of reinforcement learning, we suggest that the efficiency of this type of learning resides in the dynamic and synergistic cooperation of brain systems that use different levels of computations. The implementation of reward signals at the synaptic, cellular, network and system levels give the organism the necessary robustness, adaptability and processing speed required for evolutionary and behavioral success. © Springer+Business Media B.V. 2010.
- Stiefel, K. M., Fellous, J., Thomas, P. J., & Sejnowski, T. J. (2010). Erratum: Intrinsic subthreshold oscillations extend the influence of inhibitory synaptic inputs on cortical pyramidal neurons (European Journal of Neuroscience 31 (1019-1026)). European Journal of Neuroscience, 31(8), 1509-.
- Stiefel, K. M., Fellous, J., Thomas, P. J., & Sejnowski, T. J. (2010). Intrinsic subthreshold oscillations extend the influence of inhibitory synaptic inputs on cortical pyramidal neurons. European Journal of Neuroscience, 31(6), 1019-1026.More infoPMID: 20377616;PMCID: PMC2862239;Abstract: Fast inhibitory synaptic inputs, which cause conductance changes that typically last for 10-100 ms, participate in the generation and maintenance of cortical rhythms. We show here that these fast events can have influences that outlast the duration of the synaptic potentials by interacting with subthreshold membrane potential oscillations. Inhibitory postsynaptic potentials (IPSPs) in cortical neurons in vitro shifted the oscillatory phase for several seconds. The phase shift caused by two IPSPs or two current pulses summed non-linearly. Cholinergic neuromodulation increased the power of the oscillations and decreased the magnitude of the phase shifts. These results show that the intrinsic conductances of cortical pyramidal neurons can carry information about inhibitory inputs and can extend the integration window for synaptic input. © Federation of European Neuroscience Societies and Blackwell Publishing Ltd.
- Wang, H., Spencer, D., Fellous, J., & Sejnowski, T. J. (2010). Synchrony of Thalamocortical Inputs Maximizes Cortical Reliability. SCIENCE, 328(5974), 106-109.
- Wang, H., Spencer, D., Fellous, J., & Sejnowski, T. J. (2010). Synchrony of thalamocortical inputs maximizes cortical reliability. Science, 328(5974), 106-109.More infoPMID: 20360111;PMCID: PMC2859205;Abstract: Thalamic inputs strongly drive neurons in the primary visual cortex, even though these neurons constitute only ∼5% of the synapses on layer 4 spiny stellate simple cells. We modeled the feedforward excitatory and inhibitory inputs to these cells based on in vivo recordings in cats, and we found that the reliability of spike transmission increased steeply between 20 and 40 synchronous thalamic inputs in a time window of 5 milliseconds, when the reliability per spike was most energetically efficient. The optimal range of synchronous inputs was influenced by the balance of background excitation and inhibition in the cortex, which could gate the flow of information into the cortex. Ensuring reliable transmission by spike synchrony in small populations of neurons may be a general principle of cortical function.
- Zebrowitz, L. A., Kikuchi, M., & Fellous, J. (2010). Facial Resemblance to Emotions: Group Differences, Impression Effects, and Race Stereotypes. JOURNAL OF PERSONALITY AND SOCIAL PSYCHOLOGY, 98(2), 175-189.
- Zebrowitz, L. A., Kikuchi, M., & Fellous, J. (2010). Facial Resemblance to Emotions: Group Differences, Impression Effects, and Race Stereotypes. Journal of Personality and Social Psychology, 98(2), 175-189.More infoPMID: 20085393;PMCID: PMC3677560;Abstract: The authors used connectionist modeling to extend previous research on emotion overgeneralization effects. Study 1 demonstrated that neutral expression male faces objectively resemble angry expressions more than female faces do, female faces objectively resemble surprise expressions more than male faces do, White faces objectively resemble angry expressions more than Black or Korean faces do, and Black faces objectively resemble happy and surprise expressions more than White faces do. Study 2 demonstrated that objective resemblance to emotion expressions influences trait impressions even when statistically controlling possible confounding influences of attractiveness and babyfaceness. It further demonstrated that emotion overgeneralization is moderated by face race and that racial differences in emotion resemblance contribute to White perceivers' stereotypes of Blacks and Asians. These results suggest that intergroup relations may be strained not only by cultural stereotypes but also by adaptive responses to emotion expressions that are overgeneralized to groups whose faces subtly resemble particular emotions. © 2010 American Psychological Association.
- Fellous, J., Tatsuno, M., & Amari, S. (2009). Information-Geometric Measures as Robust Estimators of Connection Strengths and External Inputs. Neural Computation, 21(8), 2309-2335. doi:10.1162/neco.2009.04-08-748
- Phillips-portillo, J., Paulk, A. C., Gronenberg, W., Fellous, J. M., & Dacks, A. M. (2009). Visual processing in the central bee brain. The Journal of neuroscience : the official journal of the Society for Neuroscience, 29(32), 9987-99. doi:10.1523/jneurosci.1325-09.2009More infoVisual scenes comprise enormous amounts of information from which nervous systems extract behaviorally relevant cues. In most model systems, little is known about the transformation of visual information as it occurs along visual pathways. We examined how visual information is transformed physiologically as it is communicated from the eye to higher-order brain centers using bumblebees, which are known for their visual capabilities. We recorded intracellularly in vivo from 30 neurons in the central bumblebee brain (the lateral protocerebrum) and compared these neurons to 132 neurons from more distal areas along the visual pathway, namely the medulla and the lobula. In these three brain regions (medulla, lobula, and central brain), we examined correlations between the neurons' branching patterns and their responses primarily to color, but also to motion stimuli. Visual neurons projecting to the anterior central brain were generally color sensitive, while neurons projecting to the posterior central brain were predominantly motion sensitive. The temporal response properties differed significantly between these areas, with an increase in spike time precision across trials and a decrease in average reliable spiking as visual information processing progressed from the periphery to the central brain. These data suggest that neurons along the visual pathway to the central brain not only are segregated with regard to the physical features of the stimuli (e.g., color and motion), but also differ in the way they encode stimuli, possibly to allow for efficient parallel processing to occur.
- Tatsuno, M., Fellous, J., & Amari, S. (2009). Information-geometric measures as robust estimators of connection strengths and external inputs.. Neural computation, 21(8), 2309-2335.More infoPMID: 19538092;Abstract: Information geometry has been suggested to provide a powerful tool for analyzing multineuronal spike trains. Among several advantages of this approach, a significant property is the close link between information-geometric measures and neural network architectures. Previous modeling studies established that the first- and second-order information-geometric measures corresponded to the number of external inputs and the connection strengths of the network, respectively. This relationship was, however, limited to a symmetrically connected network, and the number of neurons used in the parameter estimation of the log-linear model needed to be known. Recently, simulation studies of biophysical model neurons have suggested that information geometry can estimate the relative change of connection strengths and external inputs even with asymmetric connections. Inspired by these studies, we analytically investigated the link between the information-geometric measures and the neural network structure with asymmetrically connected networks of N neurons. We focused on the information-geometric measures of orders one and two, which can be derived from the two-neuron log-linear model, because unlike higher-order measures, they can be easily estimated experimentally. Considering the equilibrium state of a network of binary model neurons that obey stochastic dynamics, we analytically showed that the corrected first- and second-order information-geometric measures provided robust and consistent approximation of the external inputs and connection strengths, respectively. These results suggest that information-geometric measures provide useful insights into the neural network architecture and that they will contribute to the study of system-level neuroscience.
- Günay, C., Smolinski, T. G., Lytton, W. W., Morse, T. M., Gleeson, P., Crook, S., Steuber, V., Silver, A., Voicu, H., Andrews, P., Bokil, H., Maniar, H., Loader, C., Mehta, S., Kleinfeld, D., Thomson, D., Mitra, P. P., Aaron, G., & Fellous, J. (2008). Computational intelligence in electrophysiology: Trends and open problems. Studies in Computational Intelligence, 122, 325-359.More infoAbstract: This chapter constitutes mini-proceedings of the Workshop on Physiology Databases and Analysis Software that was a part of the Annual Computational Neuroscience Meeting CNS*2007 that took place in July 2007 in Toronto, Canada (http://www.cnsorg.org). The main aim of the workshop was to bring together researchers interested in developing and using automated analysis tools and database systems for electrophysiological data. Selected discussed topics, including the review of some current and potential applications of Computational Intelligence (CI) in electrophysiology, database and electrophysiological data exchange platforms, languages, and formats, as well as exemplary analysis problems, are presented in this chapter. The authors hope that the chapter will be useful not only to those already involved in the field of electrophysiology, but also to CI researchers, whose interest will be sparked by its contents. © 2008 Springer-Verlag Berlin Heidelberg.
- Navratilova, Z., & Fellous, J. (2008). A biophysical model of cortical up and down states: Excitatory-inhibitory balance and H-current. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 5286 LNCS, 61-66.More infoAbstract: During slow-wave sleep, cortical neurons oscillate between up and down states. Using a computational model of cortical neurons with realistic synaptic transmission, we determined that reverberation of activity in a small network of about 40 pyramidal cells could account for the properties of up states in vivo. We found that experimentally accessible quantities such as membrane potential fluctuations, firing rates and up state durations could be used as indicators of the size of the network undergoing the up state. We also show that the H-current, together with feed-forward inhibition can act as a gating mechanism for up state initiation. © 2008 Springer Berlin Heidelberg.
- Navratilova, Z., Mcnaughton, B. L., & Fellous, J. M. (2008). Intrinsic current generated, omnidirectional phase precession and grid field scaling in toroidal attractor model of medial entorhinal path integration. BMC Neuroscience, 9(1), =. doi:10.1186/1471-2202-9-s1-p21
- Phillips-portillo, J., Paulk, A. C., Gronenberg, W., Fellous, J. M., & Dacks, A. M. (2008). The processing of color, motion, and stimulus timing are anatomically segregated in the bumblebee brain. The Journal of neuroscience : the official journal of the Society for Neuroscience, 28(25), 6319-32. doi:10.1523/jneurosci.1196-08.2008More infoAnimals use vision to perform such diverse behaviors as finding food, interacting socially with other animals, choosing a mate, and avoiding predators. These behaviors are complex and the visual system must process color, motion, and pattern cues efficiently so that animals can respond to relevant stimuli. The visual system achieves this by dividing visual information into separate pathways, but to what extent are these parallel streams separated in the brain? To answer this question, we recorded intracellularly in vivo from 105 morphologically identified neurons in the lobula, a major visual processing structure of bumblebees (Bombus impatiens). We found that these cells have anatomically segregated dendritic inputs confined to one or two of six lobula layers. Lobula neurons exhibit physiological characteristics common to their respective input layer. Cells with arborizations in layers 1-4 are generally indifferent to color but sensitive to motion, whereas layer 5-6 neurons often respond to both color and motion cues. Furthermore, the temporal characteristics of these responses differ systematically with dendritic branching pattern. Some layers are more temporally precise, whereas others are less precise but more reliable across trials. Because different layers send projections to different regions of the central brain, we hypothesize that the anatomical layers of the lobula are the structural basis for the segregation of visual information into color, motion, and stimulus timing.
- Tiesinga, P., Fellous, J., & Sejnowski, T. J. (2008). Regulation of spike timing in visual cortical circuits. NATURE REVIEWS NEUROSCIENCE, 9(2), 97-109.
- Tiesinga, P., Fellous, J., & Sejnowski, T. J. (2008). Regulation of spike timing in visual cortical circuits. Nature Reviews Neuroscience, 9(2), 97-107.More infoPMID: 18200026;PMCID: PMC2868969;Abstract: A train of action potentials (a spike train) can carry information in both the average firing rate and the pattern of spikes in the train. But can such a spike-pattern code be supported by cortical circuits? Neurons in vitro produce a spike pattern in response to the injection of a fluctuating current. However, cortical neurons in vivo are modulated by local oscillatory neuronal activity and by top-down inputs. In a cortical circuit, precise spike patterns thus reflect the interaction between internally generated activity and sensory information encoded by input spike trains. We review the evidence for precise and reliable spike timing in the cortex and discuss its computational role. © 2008 Nature Publishing Group.
- Fellous, J. M. (2007). Models of emotion.. Scholarpedia, 2(11), 1453. doi:10.4249/scholarpedia.1453
- Fellous, J. M., Corral-frias, N. S., & Buntaine, A. (2007). Emergence of reliable spike patterns in models of CA1 cells contacted by unreliable synapses. BMC Neuroscience, 8(2), -. doi:10.1186/1471-2202-8-s2-p71More infoExcitatory synapses onto CA1 pyramidal cells fail four times out of five on average, yet the firing of CA1 neuron is elicited at specific phases of the EEG theta cycle with a high degree of precision when a rat is traversing a place field. We use a multicompartmental biophysical model of several reconstructed CA1 cells, and a model of a stochastic glutamatergic synapse that includes facilitation and depression to study the conditions and properties for reliable and precise CA1 firing. The model of the synapse is tightly constrained by experimental data obtained with minimal stimulations in vitro. Synapses are presynaptically stimulated with CA3/entorhinal spike trains that have been recorded in vivo in the behaving rat. We report that under those conditions, CA1 pyramidal cells are capable of generating precise spike patterns that are theta-modulated. The precise timing of the pattern depends mainly on either the recruitment of high initial probability synapses, or on the recruitment of weaker perisomatic synapses receiving fast bursts of 1–3 presynaptic action potentials. The patterns generated are robust to noise, and contain a marked theta-frequency component, even though the input spikes are not coherent at any particular frequency. We also report that spike patterns may include gamma-like frequency components in part due to the synaptic dynamics, and to the presence of fast bursts in the presynaptic inputs. We conclude that even though afferent synapses are unreliable, CA1 pyramidal cells are able to generate precisely timed patterns of spiking that mimic those that are reported in vivo. Our model further predicts that about 400 presynaptic cells are involved in the firing of a single CA1 pyramidal cell when the rat traverses a place field.
- Fellous, J., Zebrowitz, L. A., & Kikuchi, M. (2007). Are Effects of Emotion Expression on Trait Impressions Mediated by Babyfaceness? Evidence From Connectionist Modeling. Personality and Social Psychology Bulletin, 33(5), 648-662. doi:10.1177/0146167206297399
- Sejnowski, T. J., Spencer, D., Sejnowski, T. J., Hsi-ping, W., & Fellous, J. M. (2007). Supralinear reliability of cortical spiking from synchronous thalamic inputs.. BMC Neuroscience, 8(2). doi:10.1186/1471-2202-8-s2-p130More infoThalamic and cortical V1 layer 4 neurons are capable of firing highly reliably and precisely upon repeated presentations of the same visual stimulus to the retina. To compare candidate causal mechanisms of spike-time reliability, a reconstructed multicompartment spiny stellate cell model with dynamic stochastic synapses was given varying synaptic inputs. We found reliability was primarily influenced by the number of synapses that fired synchronously during events (synchrony magnitude), which exhibits a supralinear relation; rather than by the rate of synchronous firing events (event rate) or synaptic strength, which exhibits comparatively more linear relations, even in the absence of voltage dependent conductances. Supralinear reliability highlights the efficacy of synchronous but weak synapses in driving output spiking, and may have implications for neural synchronicity within and between cortical areas.
- Zebrowitz, L. A., Kikuchi, M., & Fellous, J. (2007). Are effects of emotion expression on Trait Impressions mediated by babyfaceness? evidence from connectionist modeling. Personality and Social Psychology Bulletin, 33(5), 648-662.More infoPMID: 17440203;Abstract: Two studies provided evidence that bolsters the Marsh, Adams, and Kleck hypothesis that the morphology of certain emotion expressions reflects an evolved adaptation to mimic babies or mature adults. Study 1 found differences in emotion expressions' resemblance to babies using objective indices of babyfaceness provided by connectionist models that are impervious to overlapping cultural stereotypes about babies and the emotions. Study 2 not only replicated parallels between impressions of certain emotions and babies versus adults but also showed that objective indices of babyfaceness partially mediated impressions of the emotion expressions. babyface effects were independent of strong effects of attractiveness, and babyfaceness did not mediate impressions of happy expressions, to which the evolutionary hypothesis would not apply. © 2007 by the Society for Personality and Social Psychology, Inc.
- Mishra, J., Fellous, J., & Sejnowski, T. J. (2006). Selective attention through phase relationship of excitatory and inhibitory input synchrony in a model cortical neuron. Neural Networks, 19(9), 1329-1346.More infoPMID: 17027225;PMCID: PMC1815390;Abstract: Neurons in area V 2 and V 4 exhibit stimulus specific tuning to single stimuli, and respond at intermediate firing rates when presented with two differentially preferred stimuli ('pair response'). Selective attention to one of the two stimuli causes the neuron's firing rate to shift from the intermediate pair response towards the response to the attended stimulus as if it were presented alone. Attention to single stimuli reduces the response threshold of the neuron and increases spike synchronization at gamma frequencies. The intrinsic and network mechanisms underlying these phenomena were investigated in a multi-compartmental biophysical model of a reconstructed cat V 4 neuron. Differential stimulus preference was generated through a greater ratio of excitatory to inhibitory synapses projecting from one of two input V 2 populations. Feedforward inhibition and synaptic depression dynamics were critical to generating the intermediate pair response. Neuronal gain effects were simulated using gamma frequency range correlations in the feedforward excitatory and inhibitory inputs to the V 4 neuron. For single preferred stimulus presentations, correlations within the inhibitory population out of phase with correlations within the excitatory input significantly reduced the response threshold of the V 4 neuron. The pair response to simultaneously active preferred and non-preferred V 2 populations could also undergo an increase or decrease in gain via the same mechanism, where correlations in feedforward inhibition are out of phase with gamma band correlations within the excitatory input corresponding to the attended stimulus. The results of this model predict that top-down attention may bias the V 4 neuron's response using an inhibitory correlation phase shift mechanism. © 2006 Elsevier Ltd. All rights reserved.
- Polikov, V. S., Block, M. L., Fellous, J., Hong, J., & Reichert, W. M. (2006). In vitro model of glial scarring around neuroelectrodes chronically implanted in the CNS. Biomaterials, 27(31), 5368-5376.More infoPMID: 16842846;Abstract: A novel in vitro model of glial scarring was developed by adapting a primary cell-based system previously used for studying neuroinflammatory processes in neurodegenerative disease. Midbrains from embryonic day 14 Fischer 344 rats were mechanically dissociated and grown on poly-d-lysine coated 24 well plates to a confluent layer of neurons, astrocytes, and microglia. The culture was injured with either a mechanical scrape or foreign-body placement (segments of 50 μm diameter stainless steel microwire), fixed at time points from 6 h to 10 days, and assessed by immunocytochemistry. Microglia invaded the scraped wound area at early time points and hypertrophied activated astrocytes repopulated the wound after 7 days. The chronic presence of microwire resulted in a glial scar forming at 10 days, with microglia forming an inner layer of cells coating the microwire, while astrocytes surrounded the microglial core with a network of cellular processes containing upregulated GFAP. Vimentin expressing cells and processes were present in the scrape at early times and within the astrocyte processes forming the glial scar. Neurons within the culture did not repopulate the scrape wound and did not respond to the microwire, although they were determined to be electrically active through patch clamp recording. The time course and relative positions of the glia in response to the different injury paradigms correlated well with stereotypical in vivo responses and warrant further work in the development of a functional in vitro test bed. © 2006 Elsevier Ltd. All rights reserved.
- Robinson, J., & Fellous, J. M. (2006). A Mechanistic View of the Expression and Experience of Emotion in the Arts. American Journal of Psychology, 119(4), 668. doi:10.2307/20445371
- Bazhenov, M., Rulkov, N. F., Fellous, J., & Timofeev, I. (2005). Role of network dynamics in shaping spike timing reliability. Physical Review E - Statistical, Nonlinear, and Soft Matter Physics, 72(4).More infoPMID: 16383416;Abstract: We study the reliability of cortical neuron responses to periodically modulated synaptic stimuli. Simple map-based models of two different types of cortical neurons are constructed to replicate the intrinsic resonances of reliability found in experimental data and to explore the effects of those resonance properties on collective behavior in a cortical network model containing excitatory and inhibitory cells. We show that network interactions can enhance the frequency range of reliable responses and that the latter can be controlled by the strength of synaptic connections. The underlying dynamical mechanisms of reliability enhancement are discussed. © 2005 The American Physical Society.
- Arbib, M. A., & Fellous, J. (2004). Emotions: From brain to robot. Trends in Cognitive Sciences, 8(12), 554-561.More infoPMID: 15556025;Abstract: Some robots have been given emotional expressions in an attempt to improve human-computer interaction. In this article we analyze what it would mean for a robot to have emotion, distinguishing emotional expression for communication from emotion as a mechanism for the organization of behavior. Research on the neurobiology of emotion yields a deepening understanding of interacting brain structures and neural mechanisms rooted in neuromodulation that underlie emotions in humans and other animals. However, the chemical basis of animal function differs greatly from the mechanics and computations of current machines. We therefore abstract from biology a functional characterization of emotion that does not depend on physical substrate or evolutionary history, and is broad enough to encompass the possible emotions of robots.
- Arbib, M. A., & Fellous, J. (2004). Emotions: from brain to robot. TRENDS IN COGNITIVE SCIENCES, 8(12), 554-561.
- Fellous, J. (2004). From human emotions to robot emotions. AAAI Spring Symposium - Technical Report, 2, 37-47.More infoAbstract: The main difficulties that researchers face in understanding emotions are difficulties only because of the narrow-mindedness of our views on emotions. We are not able to free ourselves from the notion that emotions are necessarily human emotions. I will argue that if animals have emotions, then so can robots. Studies in neuroscience have shown that animal models, though having limitations, have significantly contributed to our understanding of the functional and mechanistic aspects of emotions. I will suggest that one of the main functions of emotions is to achieve the multi-level communication of simplified but high impact information. The way this function is achieved in the brain depends on the species, and on the specific emotion considered. The classical view that emotions are 'computed' by specialized brain centers, such as the 'limbic system', is criticized. I will suggest that an ensemble of well-known neurobiological phenomena, together referred to as neuromodulation, provide a useful framework for understanding how emotions arise, are maintained, and interact with other aspects of behavior and cognitive processing. This framework suggests new ways in which robot emotions can be implemented and fulfill their function. Copyright © 2004, American Association for Artificial Intelligence (www.aaai.org). All rights reserved.
- Fellous, J. M., Tiesinga, P., Thomas, P. J., & Sejnowski, T. J. (2004). Discovering spike patterns in neuronal responses. JOURNAL OF NEUROSCIENCE, 24(12), 2989-3001.
- Fellous, J., H., P., Thomas, P. J., & Sejnowski, T. J. (2004). Discovering Spike Patterns in Neuronal Responses. Journal of Neuroscience, 24(12), 2989-3001.More infoPMID: 15044538;PMCID: PMC2928855;Abstract: When a cortical neuron is repeatedly injected with the same fluctuating current stimulus (frozen noise) the timing of the spikes is highly precise from trial to trial and the spike pattern appears to be unique. We show here that the same repeated stimulus can produce more than one reliable temporal pattern of spikes. A new method is introduced to find these patterns in raw multitrial data and is tested on surrogate data sets. Using it, multiple coexisting spike patterns were discovered in pyramidal cells recorded from rat prefrontal cortex in vitro, in data obtained in vivo from the middle temporal area of the monkey (Buracas et al., 1998) and from the cat lateral geniculate nucleus (Reinagel and Reid, 2002). The spike patterns lasted from a few tens of milliseconds in vitro to several seconds in vivo. We conclude that the prestimulus history of a neuron may influence the precise timing of the spikes in response to a stimulus over a wide range of time scales.
- Schreiber, S., Fellous, J. M., Tiesinga, P., & Sejnowski, T. J. (2004). Influence of ionic conductances on spike timing reliability of cortical neurons for suprathreshold rhythmic inputs. JOURNAL OF NEUROPHYSIOLOGY, 91(1), 194-205.
- Schreiber, S., Fellous, J., Tiesinga, P., & Sejnowski, T. J. (2004). Influence of Ionic Conductances on Spike Timing Reliability of Cortical Neurons for Suprathreshold Rhythmic Inputs. Journal of Neurophysiology, 91(1), 194-205.More infoPMID: 14507985;PMCID: PMC2928819;Abstract: Spike timing reliability of neuronal responses depends on the frequency content of the input. We investigate how intrinsic properties of cortical neurons affect spike timing reliability in response to rhythmic inputs of suprathreshold mean. Analyzing reliability of conductance-based cortical model neurons on the basis of a correlation measure, we show two aspects of how ionic conductances influence spike timing reliability. First, they set the preferred frequency for spike timing reliability, which in accordance with the resonance effect of spike timing reliability is well approximated by the firing rate of a neuron in response to the DC component in the input. We demonstrate that a slow potassium current can modulate the spike timing frequency preference over a broad range of frequencies. This result is confirmed experimentally by dynamic-clamp recordings from rat prefrontal cortical neurons in vitro. Second, we provide evidence that ionic conductances also influence spike timing beyond changes in preferred frequency. Cells with the same DC firing rate exhibit more reliable spike timing at the preferred frequency and its harmonics if the slow potassium current is larger and its kinetics are faster, whereas a larger persistent sodium current impairs reliability. We predict that potassium channels are an efficient target for neuromodulators that can tune spike timing reliability to a given rhythmic input.
- Tiesinga, P. H., Fellous, J. M., Salinas, E., Jose, J. V., & Sejnowski, T. J. (2004). Inhibitory synchrony as a mechanism for attentional gain modulation. JOURNAL OF PHYSIOLOGY-PARIS, 98(4-6), 296-314.
- Tiesinga, P. H., Fellous, J., Salinas, E., José, J. V., & Sejnowski, T. J. (2004). Inhibitory synchrony as a mechanism for attentional gain modulation. Journal of Physiology Paris, 98(4-6 SPEC. ISS.), 296-314.More infoPMID: 16274973;PMCID: PMC2872773;Abstract: Recordings from area V4 of monkeys have revealed that when the focus of attention is on a visual stimulus within the receptive field of a cortical neuron, two distinct changes can occur: The firing rate of the neuron can change and there can be an increase in the coherence between spikes and the local field potential (LFP) in the gamma-frequency range (30-50 Hz). The hypothesis explored here is that these observed effects of attention could be a consequence of changes in the synchrony of local interneuron networks. We performed computer simulations of a Hodgkin-Huxley type neuron driven by a constant depolarizing current, I, representing visual stimulation and a modulatory inhibitory input representing the effects of attention via local interneuron networks. We observed that the neuron's firing rate and the coherence of its output spike train with the synaptic inputs was modulated by the degree of synchrony of the inhibitory inputs. When inhibitory synchrony increased, the coherence of spiking model neurons with the synaptic input increased, but the firing rate either increased or remained the same. The mean number of synchronous inhibitory inputs was a key determinant of the shape of the firing rate versus current (f-I) curves. For a large number of inhibitory inputs (∼50), the f-I curve saturated for large I and an increase in input synchrony resulted in a shift of sensitivity-the model neuron responded to weaker inputs I. For a small number (∼10), the f-I curves were non-saturating and an increase in input synchrony led to an increase in the gain of the response-the firing rate in response to the same input was multiplied by an approximately constant factor. The firing rate modulation with inhibitory synchrony was highest when the input network oscillated in the gamma frequency range. Thus, the observed changes in firing rate and coherence of neurons in the visual cortex could be controlled by top-down inputs that regulated the coherence in the activity of a local inhibitory network discharging at gamma frequencies. © 2005 Elsevier Ltd. All rights reserved.
- Tiesinga, P. H., Fellous, J., Salinas, E., José, J. V., & Sejnowski, T. J. (2004). Synchronization as a mechanism for attentional gain modulation. Neurocomputing, 58-60, 641-646.More infoAbstract: Responses of neurons in monkey visual cortex are modulated when attention is directed into the receptive field of the neuron: the gain or sensitivity of the response is increased or the synchronization of the spikes to the local field potential (LFP) is increased. We investigated, using model simulations, whether the synchrony of inhibitory networks could link these observations. We found that, indeed, an increase in inhibitory synchrony could enhance the coherence of the model neurons with the simulated LFP, and could have different effects on the firing rate. When the firing rate vs. current (f-I) response curves saturated at high I, attention yielded a shift in sensitivity; alternatively, when the f-I curves were non-saturating, the most significant effect was on the gain of the response. This suggests that attention may act through changes in the synchrony of inhibitory networks. © 2004 Elsevier B.V. All rights reserved.
- Fellous, J. -., Rudolph, M., Destexhe, A., & Sejnowski, T. J. (2003). Synaptic background noise controls the input/output characteristics of single cells in an in vitro model of in vivo activity. Neuroscience, 122(3), 811-829.More infoPMID: 14622924;PMCID: PMC2928821;Abstract: In vivo, in vitro and computational studies were used to investigate the impact of the synaptic background activity observed in neocortical neurons in vivo. We simulated background activity in vitro using two stochastic Ornstein-Uhlenbeck processes describing glutamatergic and GABAergic synaptic conductances, which were injected into a cell in real time using the dynamic clamp technique. With parameters chosen to mimic in vivo conditions, layer 5 rat prefrontal cortex cells recorded in vitro were depolarized by about 15 mV, their membrane fluctuated with a S.D. of about 4 mV, their input resistances decreased five-fold, their spontaneous firing had a high coefficient of variation and an average firing rate of about 5-10 Hz. Brief changes in the variance of the α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) synaptic conductance fluctuations induced time-locked spiking without significantly changing the average membrane potential of the cell. These transients mimicked increases in the correlation of excitatory inputs. Background activity was highly effective in modulating the firing-rate/current curve of the cell: the variance of the simulated γ-aminobutyric acid (GABA) and AMPA conductances individually set the input/output gain, the mean excitatory and inhibitory conductances set the working point, and the mean inhibitory conductance controlled the input resistance. An average ratio of inhibitory to excitatory mean conductances close to 4 was optimal in generating membrane potential fluctuations with high coefficients of variation. We conclude that background synaptic activity can dynamically modulate the input/output properties of individual neocortical neurons in vivo. © 2003 IBRO. Published by Elsevier Ltd. All rights reserved.
- Fellous, J. M., Rudolph, M., Destexhe, A., & Sejnowski, T. J. (2003). Synaptic background noise controls the input/output characteristics of single cells in an in vitro model of in vivo activity. NEUROSCIENCE, 122(3), 811-829.
- Fellous, J., & Sejnowski, T. J. (2003). Regulation of Persistent Activity by Background Inhibition in an In Vitro Model of a Cortical Microcircuit. Cerebral Cortex, 13(11), 1232-1241.More infoPMID: 14576214;PMCID: PMC2928820;Abstract: We combined in vitro intracellular recording from prefrontal cortical neurons with simulated synaptic activity of a layer 5 prefrontal microcircuit using a dynamic clamp. During simulated in vivo background conditions, the cell responded to a brief depolarization with a sequence of spikes that outlasted the depolarization, mimicking the activity of a cell recorded during the delay period of a working memory task in the behaving monkey. The onset of sustained activity depended on the number of action potentials elicited by the cue-like depolarization. Too few spikes failed to provide enough NMDA drive to elicit sustained reverberations; too many spikes activated a slow intrinsic hyperpolarization current that prevented spiking; an intermediate number of spikes produced sustained activity. When high dopamine levels were simulated by depolarizing the cell and by increasing the amount of NMDA current, the cell exhibited spontaneous 'up-states' that terminated by the activation of a slow a intrinsic hyperpolarizing current. The firing rate during the delay period could be effectively modulated by the standard deviation of the inhibitory background synaptic noise without significant changes in the background firing rate before cue onset. These results suggest that the balance between fast feedback inhibition and slower AMPA and NMDA feedback excitation is critical in initiating persistent activity and that the maintenance of persistent activity may be regulated by the amount of correlated background inhibition.
- Schreiber, S., Fellous, J. M., Whitmer, D., Tiesinga, P., & Sejnowski, T. J. (2003). A new correlation-based measure of spike timing reliability. NEUROCOMPUTING, 52-4, 925-931.
- Schreiber, S., Fellous, J. M., Whitmer, D., Tiesinga, P., & Sejnowski, T. J. (2003). A new correlation-based measure of spike timing reliability. Neurocomputing, 52-54, 925-931.More infoAbstract: We introduce a new correlation-based measure of spike timing reliability. Unlike other measures, it does not require the definition of a posteriori "events". It relies on only one parameter, which relates to the timescale of spike timing precision. We test the measure on surrogate data sets with varying amounts of spike time jitter, and missing or additional spikes, and compare it with a widely used histogram-based measure. The measure is efficient and faithful in characterizing spike timing reliability and produces smaller errors in the reliability estimate than the histogram-based measure based on the same number of trials. © 2002 Elsevier Science B.V. All rights reserved.
- Thomas, P. J., Tiesinga, P. H., Fellous, J., & Sejnowski, T. J. (2003). Reliability and bifurcation in neurons driven by multiple sinusoids. Neurocomputing, 52-54, 955-961.More infoAbstract: The brain produces dynamical rhythms at many frequencies that shift in amplitude and phase. To understand the functional consequences of mixtures of oscillations at the single cell level, we recorded the spike trains from single rat cortical neurons in vitro in response to two mixed sine wave currents. The reliability of spike timing was measured as a function of the relative power, phase and frequencies of the sine wave mixture. Peaks in the reliability were observed at a preferred phase difference, frequency and relative power. These results have a natural interpretation in terms of spike train attractors and bifurcations. © 2003 Elsevier Science B.V. All rights reserved.
- Zebrowitz, L. A., Fellous, J. M., Mignault, A., & Andreoletti, C. (2003). Trait impressions as overgeneralized responses to adaptively significant facial qualities: Evidence from connectionist modeling. PERSONALITY AND SOCIAL PSYCHOLOGY REVIEW, 7(3), 194-215.
- Zebrowitz, L. A., Fellous, J., Mignault, A., & Andreoletti, C. (2003). Trait impressions as overgeneralized responses to adaptively significant facial qualities: Evidence from connectionist modeling. Personality and Social Psychology Review, 7(3), 194-215.More infoPMID: 12788687;Abstract: Connectionist modeling experiments tested anomalous-face and baby-face overgeneralization hypotheses proposed to explain consensual trait impressions of faces. Activation of a neural network unit trained to respond to anomalous faces predicted impressions of normal adult faces varying in attractiveness as well as several elderly stereotypes. Activation of a neural network unit trained to respond to babies' faces predicted impressions of adults varying in babyfaceness as well as 1 elderly stereotype. Thus, similarities of normal adult faces to anomalous faces or babies' faces contribute to impressions of them quite apart from knowledge of overlapping social stereotypes. The evolutionary importance of appropriate responses to unfit individuals or babies is presumed to produce a strong response preparedness that is overgeneralized to faces resembling the unfit or babies.
- Tiesinga, P. H., Fellous, J. -., & Sejnowski, T. J. (2002). Attractor reliability reveals deterministic structure in neuronal spike trains. Neural Computation, 14(7), 1629-1650.More infoPMID: 12079549;Abstract: When periodic current is injected into an integrate-and-fire model neuron, the voltage as a function of time converges from different initial conditions to an attractor that produces reproducible sequences of spikes. The attractor reliability is a measure of the stability of spike trains against intrinsic noise and is quantified here as the inverse of the number of distinct spike trains obtained in response to repeated presentations of the same stimulus. High reliability characterizes neurons that can support a spike-time code, unlike neurons with discharges forming a renewal process (such as a Poisson process). These two classes of responses cannot be distinguished using measures based on the spike-time histogram, but they can be identified by the attractor dynamics of spike trains, as shown here using a new method for calculating the attractor reliability. We applied these methods to spike trains obtained from current injection into cortical neurons recorded in vitro. These spike trains did not form a renewal process and had a higher reliability compared to renewal-like processes with the same spike-time histogram.
- Tiesinga, P. H., Fellous, J. -., & Sejnowski, T. J. (2002). Spike-time reliability of periodically driven integrate-and-fire neurons. Neurocomputing, 44-46, 195-200.More infoAbstract: The response of model neurons driven by a periodic current converges onto mode-locked attractors. Reliability, defined as the noise stability of the attractor, was studied as a function of the drive frequency and noise strength. For weak noise, the neuron remained on one attractor and reliability was high. For intermediate noise strength, transitions between attractors occurred. For strong noise, mode locking became unstable. The attractor was most stable for frequencies for which the neuron produced one spike on each cycle. The prediction of a reliability resonance as a function of drive frequency has been observed in vitro in cortical neurons. © 2002 Published by Elsevier Science B.V.
- Tiesinga, P. H., Fellous, J. -., José, J., & Sejnowski, T. J. (2002). Information transfer in entrained cortical neurons. Network: Computation in Neural Systems, 13(1), 41-66.More infoAbstract: Cortical interneurons connected by gap junctions can provide a synchronized inhibitory drive that can entrain pyramidal cells. This was studied in a single-compartment Hodgkin-Huxley-type model neuron that was entrained by periodic inhibitory inputs with low jitter in the input spike times (i.e. high precision), and a variable but large number of presynaptic spikes on each cycle. During entrainment the Shannon entropy of the output spike times was reduced sharply compared with its value outside entrainment. Surprisingly, however, the information transfer as measured by the mutual information between the number of inhibitory inputs in a cycle and the phase lag of the subsequent output spike was significantly increased during entrainment. This increase was due to the reduced contribution of the internal correlations to the output variability. These theoretical predictions were supported by experimental recordings from the rat neocortex and hippocampus in vitro.
- Tiesinga, P. H., Fellous, J. M., José, J., & Sejnowski, T. J. (2002). Information transfer in entrained cortical neurons.. Network (Bristol, England), 13(1), 41-66.More infoPMID: 11878284;Abstract: Cortical interneurons connected by gap junctions can provide a synchronized inhibitory drive that can entrain pyramidal cells. This was studied in a single-compartment Hodgkin-Huxley-type model neuron that was entrained by periodic inhibitory inputs with low jitter in the input spike times (i.e. high precision), and a variable but large number of presynaptic spikes on each cycle. During entrainment the Shannon entropy of the output spike times was reduced sharply compared with its value outside entrainment. Surprisingly, however, the information transfer as measured by the mutual information between the number of inhibitory inputs in a cycle and the phase lag of the subsequent output spike was significantly increased during entrainment. This increase was due to the reduced contribution of the internal correlations to the output variability. These theoretical predictions were supported by experimental recordings from the rat neocortex and hippocampus in vitro.
- Blanchard, C., Blanchard, R., Fellous, J. -., Guimaraes, F. S., Irwin, W., LeDoux, J. E., McGaugh, J. L., Rosen, J. B., Schenberg, L. C., Volchan, E., & Cunha, C. D. (2001). The brain decade in debate: III. Neurobiology of emotion. Brazilian Journal of Medical and Biological Research, 34(3), 283-293.More infoPMID: 11262578;Abstract: This article is a transcription of an electronic symposium in which active researchers were invited by the Brazilian Society of Neuroscience and Behavior (SBNeC) to discuss the advances of the last decade in the neurobiology of emotion. Four basic questions were debated: 1) What are the most critical issues/questions in the neurobiology of emotion? 2) What do we know for certain about brain processes involved in emotion and what is controversial? 3) What kinds of research are needed to resolve these controversial issues? 4) What is the relationship between learning, memory and emotion? The focus was on the existence of different neural systems for different emotions and the nature of the neural coding for the emotional states. Is emotion the result of the interaction of different brain regions such as the amygdala, the nucleus accumbens, or the periaqueductal gray matter or is it an emergent property of the whole brain neural network? The relationship between unlearned and learned emotions was also discussed. Are the circuits of the former the underpinnings of the latter? It was pointed out that much of what we know about emotions refers to aversively motivated behaviors, like fear and anxiety. Appetitive emotions should attract much interest in the future. The learning and memory relationship with emotions was also discussed in terms of conditioned and unconditioned stimuli, innate and learned fear, contextual cues inducing emotional states, implicit memory and the property of using this term for animal memories. In a general way it could be said that learning modifies the neural circuits through which emotional responses are expressed.
- Destexhe, A., Rudolph, M., Fellous, J. -., & Sejnowski, T. J. (2001). Fluctuating synaptic conductances recreate in vivo-like activity in neocortical neurons. Neuroscience, 107(1), 13-24.More infoPMID: 11744242;PMCID: PMC3320220;Abstract: To investigate the basis of the fluctuating activity present in neocortical neurons in vivo, we have combined computational models with whole-cell recordings using the dynamic-clamp technique. A simplified 'point-conductance' model was used to represent the currents generated by thousands of stochastically releasing synapses. Synaptic activity was represented by two independent fast glutamatergic and GABAergic conductances described by stochastic random-walk processes. An advantage of this approach is that all the model parameters can be determined from voltage-clamp experiments. We show that the point-conductance model captures the amplitude and spectral characteristics of the synaptic conductances during background activity. To determine if it can recreate in vivo-like activity, we injected this point-conductance model into a single-compartment model, or in rat prefrontal cortical neurons in vitro using dynamic clamp. This procedure successfully recreated several properties of neurons intracellularly recorded in vivo, such as a depolarized membrane potential, the presence of high-amplitude membrane potential fluctuations, a low-input resistance and irregular spontaneous firing activity. In addition, the point-conductance model could simulate the enhancement of responsiveness due to background activity. We conclude that many of the characteristics of cortical neurons in vivo can be explained by fast glutamatergic and GABAergic conductances varying stochastically. © 2001 IBRO. Published by Elsevier Sciene Ltd. All rights reserved.
- Destexhe, A., Rudolph, M., Fellous, J. M., & Sejnowski, T. J. (2001). Fluctuating synaptic conductances recreate in vivo-like activity in neocortical neurons. NEUROSCIENCE, 107(1), 13-24.
- Fellous, J. -., Houweling, A. R., Modi, R. H., Rao, R. P., Tiesinga, P. H., & Sejnowski, T. J. (2001). Frequency dependence of spike timing reliability in cortical pyramidal cells and interneurons. Journal of Neurophysiology, 85(4), 1782-1787.More infoPMID: 11287500;Abstract: Pyramidal cells and interneurons in rat prefrontal cortical slices exhibit subthreshold oscillations when depolarized by constant current injection. For both types of neurons, the frequencies of these oscillations for current injection just below spike threshold were 2-10 Hz. Above spike threshold, however, the subthreshold oscillations in pyramidal cells remained low, but the frequency of oscillations in interneurons increased up to 50 Hz. To explore the interaction between these intrinsic oscillations and external inputs, the reliability of spiking in these cortical neurons was studied with sinusoidal current injection over a range of frequencies above and below the intrinsic frequency. Conical neurons produced 1:1 phase locking for a limited range of driving frequencies for fixed amplitude. For low-input amplitude, 1:1 phase locking was obtained in the 5- to 10-Hz range. For higher-input amplitudes, pyramidal cells phase-locked in the 5- to 20-Hz range, whereas interneurons phase-locked in the 5- to 50-Hz range. For the amplitudes studied here, spike time reliability was always highest during 1:1 phase-locking, between 5 and 20 Hz for pyramidal cells and between 5 and 50 Hz for interneurons. The observed differences in the intrinsic frequency preference between pyramidal cells and interneurons have implications for rhythmogenesis and information transmission between populations of cortical neurons.
- Fellous, J. -., Tiesinga, P. H., José, J., & Sejnowski, T. J. (2001). Computational model of carbachol-induced δ, θ and γ-like oscillations in hippocampus. Neurocomputing, 38-40, 587-593.More infoAbstract: Application of carbachol (cch) can induce oscillations in the δ (cch-δ, 0.5-2 Hz), θ (cch-θ, 4-12 Hz), and γ (cch-γ, 30-80 Hz) frequency-range in the rat hippocampal slice preparation. Using model CA3 cells we found that the time scale for cch-δ was determined by the decay constant of IK-AHP, that of cch-θ by an intrinsic subthreshold membrane potential oscillation, and that of cch-γ by the decay constant of GABA-ergic postsynaptic potentials. The known physiological effects of carbachol on IM, and IK-AHP, and on the strength of excitatory postsynaptic potentials produced the observed transitions from incoherent cch-θ to cch-δ, and from cch-δ to cch-θ with increasing cch concentrations, as well as the nested cch-γ-δ and cch-γ-θ seen in experiment. © 2001 Elsevier Science B.V. All rights reserved.
- Fellous, J. M., Houweling, A. R., Modi, R. H., Rao, R., Tiesinga, P., & Sejnowski, T. J. (2001). Frequency dependence of spike timing reliability in cortical pyramidal cells and interneurons. JOURNAL OF NEUROPHYSIOLOGY, 85(4), 1782-1787.
- H., P., H., P., Fellous, J., Fellous, J., José, J. V., José, J. V., Sejnowski, T. J., & Sejnowski, T. J. (2001). Computational model of carbachol-induced delta, theta, and gamma oscillations in the hippocampus. Hippocampus, 11(3), 251-274.More infoPMID: 11769308;Abstract: Field potential recordings from the rat hippocampus in vivo contain distinct frequency bands of activity, including δ (0.5-2 Hz), θ (4-12 Hz), and γ (30-80 Hz), that are correlated with the behavioral state of the animal. The cholinergic agonist carbachol (CCH) induces oscillations in the δ (CCH-δ), θ (CCH-θ), and γ (CCH-γ) frequency ranges in the hippocampal slice preparation, eliciting asynchronous CCH-θ, synchronous CCH-δ, and synchronous CCH-θ with increasing CCH concentration (Fellous and Sejnowski, Hippocampus 2000;10:187-197). In a network model of area CA3, the time scale for CCH-δ corresponded to the decay constant of the gating variable of the calcium-dependent potassium (K-AHP) current, that of CCH-θ to an intrinsic subthreshold membrane potential oscillation of the pyramidal cells, and that of CCH-γ to the decay constant of GABAergic inhibitory synaptic potentials onto the pyramidal cells. In model simulations, the known physiological effects of carbachol on the muscarinic and K-AHP currents, and on the strengths of excitatory postsynaptic potentials, reproduced transitions from asynchronous CCH-θ to CCH-δ and from CCH-δ to synchronous CCH-θ. The simulations also exhibited the interspersed CCH-γ/CCH-δ and CCH-γ/CCH-θ that were observed in experiments. The model, in addition, predicted an oscillatory state with all three frequency bands present, which has not yet been observed experimentally. © 2001 Wiley-Liss, Inc.
- Houweling, A. R., Modi, R. H., Ganter, P., Fellous, J., & Sejnowski, T. J. (2001). Models of frequency preferences of prefrontal cortical neurons. Neurocomputing, 38-40, 231-238.More infoAbstract: The reliability of spike trains generated by sinusoidal current injections in prefrontal cortical pyramidal cells and interneurons depends strongly on the input frequency. We constructed computational models in order to study how cellular properties affect reliability. The models reproduced the main experimental findings: subthreshold oscillations, resonance and reliability of spike timing. The amplitude of intrinsic noise in the model determined the number of reliable frequency bands. In addition, the frequency content of the noise did not affect reliability. © 2001 Elsevier Science B.V. All rights reserved.
- Scheler, G., & Fellous, J. (2001). Dopamine modulation of prefrontal delay activity-reverberatory activity and sharpness of tuning curves. Neurocomputing, 38-40, 1549-1556.More infoAbstract: Recent electrophysiological experiments have shown that dopamine (D1) modulation of pyramidal cells in prefrontal cortex reduces spike frequency adaptation and enhances NMDA transmission. Using four models, from multicompartmental to integrate-and-fire, we examine the effects of these modulations on sustained (delay) activity in a reverberatory network. We find that D1 modulation may enable robust network bistability yielding selective reverberation among cells that code for a particular item or location. We further show that the tuning curve of such cells is sharpened, and that signal-to-noise ratio is increased. We postulate that D1 modulation affects the tuning of "memory fields" and yield efficient distributed dynamic representations. © 2001 Elsevier Science B.V. All rights reserved.
- Tiesinga, P. H., Fellous, J. -., José, J., & Sejnowski, T. J. (2001). Optimal information transfer in synchronized neocortical neurons. Neurocomputing, 38-40, 397-402.More infoAbstract: The output precision and information transmission was studied in a model neocortical neuron that was driven by a periodic presynaptic spike train with a variable number of inhibitory inputs on each cycle. Spike-timing precision was maintained during feedforward propagation during entrainment. The range of presynaptic firing rates and precision for entrainment was determined. During entrainment the Shannon information of the output spike phase was reduced but the amount of information the neuron transmitted about the synaptic input was increased. We quantify how robust information transmission is against intrinsic neuronal noise. We propose how neuromodulation, via entrainment, can regulate the information transfer in neocortical networks. © 2001 Elsevier Science B.V. All rights reserved.
- Tiesinga, P., Fellous, J. M., Jose, J. V., & Sejnowski, T. J. (2001). Computational model of carbachol-induced delta, theta, and gamma oscillations in the hippocampus. HIPPOCAMPUS, 11(3), 251-274.
- Fellous, J. M., & Sejnowski, T. J. (2000). Cholinergic induction of oscillations in the hippocampal slice in the slow (0.5-2 Hz), theta (5-12 Hz), and gamma (35-70 Hz) bands. HIPPOCAMPUS, 10(2), 187-197.
- Fellous, J., & Sejnowski, T. J. (2000). Cholinergic induction of oscillations in the hippocampal slice in the slow (0.5-2 Hz), theta (5-12 Hz), and gamma (35-70 Hz) bands. Hippocampus, 10(2), 187-197.More infoPMID: 10791841;Abstract: Carbachol, a muscarinic receptor agonist, produced three distinct spontaneous oscillations in the CA3 region of rat hippocampal slices. Carbachol concentrations in the 4-13 μM range produced regular synchronized CA3 discharges at 0.5-2 Hz (carbachol-delta). Higher concentrations (13-60 μM) produced short episodes of 5-10 Hz (carbachol-theta) oscillations separated by nonsynchronous activity. Concentrations of carbachol ranging from 8-25 μM also produced irregular episodes of high-frequency discharges (carbachol-gamma, 35-70 Hz), in isolation or mixed with carbachol-theta and carbachol-delta. At carbachol concentrations sufficient to induce carbachol-theta, low concentrations of APV reversibly transformed carbachol-theta into carbachol-delta. Higher concentrations of D,L-2-amino-5-phosphonopentanoic acid (APV) reversibly and completely blocked carbachol-theta. A systematic study of the effects of carbachol shows that the frequency of spontaneous oscillations depended nonlinearly on the level of muscarinic activation. Field and intracellular recordings from CA1 and CA3 pyramidal cells and interneurons during carbachol-induced rhythms revealed that the hippocampal circuitry preserved in the slice was capable of spontaneous activity over the range of frequencies observed in vivo and suggests that the presence of these rhythms could be under neuromodulatory control. (C) 2000 Wiley-Liss, Inc.
- Fellous, J. (1999). Neuromodulatory basis of emotion. Neuroscientist, 5(5), 283-294.More infoAbstract: The neural basis of emotion can be found in both the neural computation and the neuromodulation of the neural substrate that mediates behavior. I review the experimental evidence showing the involvement of the hypothalamus, the amygdala, and the prefrontal cortex in emotion. For each of these structures, I show the important role of various neuromodulatory systems in mediating emotional behavior. Generalizing, I suggest that behavioral complexity is caused partly by the diversity and intensity of neuromodulation and hence depends on emotional contexts. Rooting the emotional state in neuromodulatory phenomena allows for its quantitative and scientific study and possibly its characterization.
- Fellous, J., & Linster, C. (1998). Computational Models of Neuromodulation. Neural Computation, 10(4), 771-805.More infoPMID: 9573404;Abstract: Computational modeling of neural substrates provides an excellent theoretical framework for the understanding of the computational roles of neuromodulation. In this review, we illustrate, with a large number of modeling studies, the specific computations performed by neuromodulation in the context of various neural models of invertebrate and vertebrate preparations. We base our characterization of neuromodulations on their computational and functional roles rather than on anatomical or chemical criteria. We review the main framework in which neuromodulation has been studied theoretically (central pattern generation and oscillations, sensory processing, memory and information integration). Finally, we present a detailed mathematical overview of how neuromodulation has been implemented at the single cell and network levels in modeling studies. Overall, neuromodulation is found to increase and control computational complexity.
- Lisman, J. E., Fellous, J. M., & Wang, X. J. (1998). A role for NMDA-receptor channels in working memory. NATURE NEUROSCIENCE, 1(4), 273-275.
- Lisman, J. E., Fellous, J., & Wang, X. (1998). A role for NMDA-receptor channels in working memory. Nature Neuroscience, 1(4), 273-275.More infoPMID: 10195158;
- Fellous, J. (1997). Gender discrimination and prediction on the basis of facial metric information. Vision Research, 37(14), 1961-1973.More infoPMID: 9274781;Abstract: Horizontal and vertical facial measurements are statistically independent. Discriminant analysis shows that five of such normalized distances explain over 95% of the gender differences of 'training' samples and predict the gender of 90% novel test faces exhibiting various facial expressions. The robustness of the method and its results are assessed. It is argued that these distances (termed fiducial) are compatible with those found experimentally by psychophysical and neurophysiological studies. In consequence, partial explanations for the effects observed in these experiments can be found in the intrinsic statistical nature of the facial stimuli used.
- Wiskott, L., Fellous, J. M., Kruger, N., & vonderMalsburg, C. (1997). Face recognition by elastic bunch graph matching. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 19(7), 775-779.
- Wiskott, L., Fellous, J., Krueger, N., & von, C. (1997). Face recognition by elastic bunch graph matching. IEEE International Conference on Image Processing, 1, 129-132.More infoAbstract: We present a system for recognizing human faces from single images out of a large database containing one image per person. Faces are represented by labeled graphs, based on a Gabor wavelet transform. Image graphs of new faces are extracted by an elastic graph matching process and can be compared by a simple similarity function. The system differs from the preceding one in three respects. Phase information is used for accurate node positioning. Object-adapted graphs are used to handle large rotations in depth. Image graph extraction is based on a novel data structure, the bunch graph, which is constructed from a small set of sample image graphs.
- Wiskott, L., Fellous, J., Krüger, N., & Der, C. (1997). Face recognition by elastic bunch graph matching. IEEE Transactions on Pattern Analysis and Machine Intelligence, 19(7), 775-779.More infoAbstract: (http://www.neuroinformatik.ruhr-uni-bochum.de), when this research was performed. He is now at the Computational Neurobiology Laboratory, The Salk Institute for Biological Studies, San Diego, CA 92186-5800. E-mail: wiskott@cnl.salk.edu. • J.-M. Fellous was with the Computer Science Department, University of Southern California, Los Angeles, CA 90089 when this research was performed. He is now at the Volen Center for Complex Systems, Brandeis University, Waltham, MA 02254-9110. E-mail: fellous@cajal.ccs.brandeis.edu. • N. Krüger and C. von der Malsburg are with the Institute for Neural Computation, Bochum. Christoph von der Malsburg is also with the Computer Science Department, University of Southern California, Los Angeles. Email: {nkrueger, malsburgl@neuroinformatik.ruhr-uni-bochum.de. We present a system for recognizing human faces from single images out of a large database containing one image per person. Faces are represented by labeled graphs, based on a Gabor wavelet transform. Image graphs of new faces are extracted by an elastic graph matching process and can be compared by a simple similarity function. The system differs from the preceding one [1] in three respects. Phase information is used for accurate node positioning. Object-adapted graphs are used to handle large rotations in depth. Image graph extraction is based on a novel data structure, the bunch graph, which is constructed from a small set of sample image graphs. ©1997 IEEE.
Proceedings Publications
- Weitzenfeld, A., Scleidorovich, P., Llofriu, M., & Fellous, J. M. (2020). A Computational Model for Latent Learning based on Hippocampal Replay. In 2020 International Joint Conference on Neural Networks (IJCNN), 1-8.More infoWe show how hippocampal replay could explain latent learning, a phenomenon observed in animals where unrewarded pre-exposure to an environment, i.e. habituation, improves task learning rates once rewarded trials begin. We first describe a computational model for spatial navigation inspired by rat studies. The model exploits offline replay of trajectories previously learned by applying reinforcement learning. Then, to assess our hypothesis, the model is evaluated in a "multiple T-maze" environment where rats need to learn a path from the start of the maze to the goal. Simulation results support our hypothesis that pre-exposed or habituated rats learn the task significantly faster than non-pre-exposed rats. Results also show that this effect increases with the number of pre-exposed trials.
- Weitzenfeld, A., Tejera, G., Scleidorovich, P., Pelc, T., Llofriu, M., Fellous, J. M., & Contreras, M. (2019). A Computational Model for a Multi-Goal Spatial Navigation Task inspired by Rodent Studies. In 2019 International Joint Conference on Neural Networks (IJCNN).More infoWe present a biologically-inspired computational model of the rodent hippocampus based on recent studies of the hippocampus showing that its longitudinal axis is involved in complex spatial navigation. While both poles of the hippocampus, i.e. septal (dorsal) and temporal (ventral), encode spatial information; the septal area has traditionally been attributed more to navigation and action selection; whereas the temporal pole has been more involved with learning and motivation. In this work we hypothesize that the septal-temporal organization of the hippocampus axis also provides a multi-scale spatial representation that may be exploited during complex rodent navigation. To test this hypothesis, we developed a multi-scale model of the hippocampus evaluated it with a simulated rat on a multi-goal task, initially in a simplified environment, and then on a more complex environment where multiple obstacles are introduced. In addition to the hippocampus providing a spatial representation of the environment, the model includes an actor-critic framework for the motivated learning of the different tasks.
- Lester, A. W., Howard, M. D., Fellous, J., Bhattacharyya, R., & , . (2013, 2013). A Computational Model of Perirhinal Cortex: Gating and Repair of Input to the Hippocampus. In 2013 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN).
- Howard, M. D., Bhattacharyya, R., O'reilly, R. C., Ascoli, G. A., & Fellous, J. (2011). Adaptive Recall in Hippocampus.. In SFN.
- Fellous, J., Suri, R. E., & Fellous, J. M. (2002). A Roles of Dopamine. In SFN.More infoINTRODUCTION Dopamine (DA) is a neuromodulator (see: NEUROMODULATION IN INVERTEBRATE NERVOUS SYSTEMS and SYNAPTIC CURRENTS, NEUROMODULATION AND KINETIC MODELS) that originates from small groups of neurons in the mesencephalon (the ventral tegmental area (A10), the substantia nigra (A9) and A8) and in the diencephalon (area A13, A14 and A15). Dopaminergic projections are in general very diffuse and reach large portions of the brain. The time scales of dopamine actions are diverse from few hundreds of milliseconds to several hours. We will focus here on the mesencephalic dopamine centers because they are the most studied, and because they are thought to be involved in diseases such as Tourette’s syndrome, schizophrenia, Parkinson’s disease, Huntington’s disease, drug addiction or depression (see DISEASE: NEURAL NETWORK MODELS and (Tzschentke, 2001)). These centers are also involved in normal brain functions such as working memory, reinforcement learning, and attention. This article briefly summarizes the main roles of dopamine in particular with respect to recent modeling approaches.
- Wiskott, L., Fellous, J., Kruger, N., Malsburg, C. V., Fellous, J. M., & Malsburg, C. V. (1999). Face recognition by elastic bunch graph matching. In sfn.
- Ledoux, J. E., Fellous, J., & Fellous, J. M. (1998, -). Emotion and computational neuroscience. In ScholarPedia, -, -.
- Wiskott, L., Malsburg, C. V., Kruger, N., & Fellous, J. M. (1997, -). Face Recognition by Elastic Bunch Graph Matching. In sfn, -, 456-463.More infoWe present a system for recognizing human faces from single images out of a large database with one image per person. The task is difficult because of image variation in terms of position, size, expression, and pose. The system collapses most of this variance by extracting concise face descriptions in the form of image graphs. In these, fiducial points on the face (eyes, mouth etc.) are described by sets of wavelet components (jets). Image graph extraction is based on a novel approach, the bunch graph, which is constructed from a small set of sample image graphs. Recognition is based on a straight-forward comparison of image graphs. We report recognition experiments on the FERET database and the Bochum database, including recognition across pose.
Presentations
- Greene, P., Lin, K., & Fellous, J. (2015, 11). A Bayesian Approach to Source Localization with Applications to Spike Sorting. Statistical Analysis of Neural Data workshop. Pittsburgh, PA.
Poster Presentations
- Fellous, J., Scleidorovich, P., & Weitzenfeld, A. (2019, Oct). A Computational Model Combining Dorsal and Ventral Hippocampal Place Field Maps: An Analysis of Multiple-Scale Contributions. Society for Neuroscience. Chicago, IL.
- Llofriu, M., Scleidorovich, P., Tejera, G., Contreras, M., Pelc, T., Fellous, J., & Weitzenfeld, A. (2019, jul). Computational Model for a Multi-Goal Spatial Navigation Task inspired by Rodent Studies. International Joint Conference on Neural Networks. Budapest, Hungary.
- Souder, M., Rogers, M., Qin, Y., Scleidorovich, P., Weitzenfeld, A., & Fellous, J. (2019, Oct). The role of objects during complex spatial navigation in the rat. Society for Neuroscience. Chicago, IL.
- Xiao, Z., Nagl, S., Lin, K., & Fellous, J. (2019, oct). Continuous reward-place coding properties of dorsal distal CA1 hippocampus cells. Society for Neuroscience. Chicago, IL.
- Fellous, J., & Contreras, M. (2018, Oct). Involvement of the anterior insular cortex in empathic response in rats. Society for Neuroscience. San-Diego, CA.
- Harland, B., Contreras, M., Scleidorovich, P., Weitzenfeld, A., & Fellous, J. (2018, Oct). Dorsal-Ventral place cell representations in multi-scale environments. Society for Neuroscience. San-Diego, CA.
- Harper, B., & Fellous, J. (2018, Oct). A method for the precise detection and validation of spindle timing in rodents. Society for Neuroscience. San-Diego, CA.
- Howard, E., Contreras, M., Harper, B., Armstrong, E., Padgett, R., & Fellous, J. (2018, Oct). Targeted memory reactivation during sleep facilitates spatial memory consolidation in rats. Society for Neuroscience. San-Diego, CA.
- Scleidorovich, P., Harland, B., Fellous, J., & Weitzenfeld, A. (2018, Oct). Modeling of Multi-Scale Spatial Navigation in Complex Environments. Society for Neuroscience. San-Diego, CA.
- Contreras, M., Hatfield, A., Cummings, J., Cruz, K., & Fellous, J. (2017, 11). Understanding the neural basis of empathy in rodents. Society for Neuroscience. Washington DC.
- Greene, P., Fellous, J., & Lin, K. (2017, 11). Spike sorting via source localization. Society for Neuroscience. Washington DC.
- Harland, B., Contreras, M., Scleidorovich, P., Llofriu, M., Cazin, N., Weitzenfeld, A., Dominey, P., & Fellous, J. (2017, 11). Complex rodent spatial navigation optimization in a large-scale environment. Society for Neuroscience. Washington DC.
- Harper, B., Contreras, M., & Fellous, J. (2017, 11). Effects of learning on the co-occurrence of hippocampal sharp-wave ripples and prefrontal cortical spindles in the rat. Society for Neuroscience. Washington DC.
- Malerba, P., Nagl, S., Fellous, J., & Bazhenov, M. (2017, 11). Reactivation of interfering memories in the hippocampus shapes memory performance: a computational study. Society for Neuroscience. Washington DC.
- Scleidorovich, P., Cazin, N., Llofriu, M., Harland, B., Fellous, J., Dominey, P., & Weitzenfeld, A. (2017, 6). An Integrated Hippocampus-Prefrontal Cortex Model for Spatial Sequence Learning by Concatenating Replayed Place-Cell Snippets. CRCNS. Providence RI.
- Scleidorovich,, P., Llofriu, M., Cazin, N., Harland, B., Dominey, P., Fellous, J., & Weitzenfeld, A. (2017, 11). Topological map learning during pre-exposure and replay as an explanation of latent learning. Society for Neuroscience. Washington DC.
- Cazin, N., Scleidorovich, P., Llofriu, M., Pelc, T., Fellous, J., Weitzenfeld, A., & Dominey, P. (2016, November). Prefrontal cortex reservoir network learns to create novel efficient navigation sequences by concatenating place-cell snippets replayed in hippocampus. Society for Neuroscience. San Diego, CA.
- Contreras, M., Cruz, K., Cummings, J., Hatfield, A., & Fellous, J. (2016, November). Exploring mechanisms for empathy-like behavior in rats. Society for Neuroscience. San Diego, 2016.
- Harper, B., Sampson, A., Sejnowski, T., & Fellous, J. (2016, November). Sleep spindles and single-cell reactivation in the rodent medial prefrontal cortex during context-dependent memory reconsolidation. Society for Neuroscience. San Diego, CA.
- Nagl, S., Harper, B., Malerba, P., Bazhenov, M., & Fellous, J. (2016, November). Neurophysiological Correlates of the Influences of Spatial Context on Hippocampal Reactivation in a Rodent Reconsolidation Paradigm. Society for Neuroscience. San Diegi, CA.
- Pelc, T., Llofriu, M., Cazin, N., Scleidorovich, P., Dominey, P., Weitzenfeld, A., & Fellous, J. (2016, November). Neurophysiological Correlates of Spatial Navigation Optimization in the Rodent. Society for Neuroscience. SanDiego, CA.
- Ragone, M., Gianelli, S., Schwartz, D., Su, L., Koyluoglu, O. O., & Fellous, J. (2016, November). The role of hippocampal replay in a computational model of path learning. Society for Neuroscience. San Diego, CA.
- Scleidorovich, P., Llofriu, M., Pelc, T., Cazin, N., Dominey, P., Fellous, J., & Weitzenfeld, A. (2016, November). Intrahippocampal cell synapse learning may support pre-exposure based latent learning by means of improved replay events. Society for Neuroscience. San Diego, CA.
- Aykin, I., Koyluoglu, O., & Fellous, J. (2015, 11). Formation of dorso-ventral grid cell modules: The role of learning. Computational and Systems Neuroscience, COSYNE. Salt Lake City, Utah.
- Cazin, N., Fellous, J., Weitzenfeld, A., & Dominey, P. (2015, 11). Prefrontal Cortex Reservoir Network Learns to Reconstruct Navigation Sequences by Concatenating Place-cell Snippets Replayed in Hippocampus. Society for Neuroscience.
- Contreras, M., Pelc, T., Llofriu, M., Weitzenfeld, A., & Fellous, J. (2015, 11). Ventral hippocampus inactivation impairs goal-directed spatial navigation in obstacle-laden environment. Society for Neuroscience.
- Llofriu, M., Calderon, J., Lopez, G., Fellous, J., & Weitzenfeld, A. (2015, 11). Bio-Inspired Multi-Scale Representation for Navigation Learning. IEEE International Conference on Robotics and Automation. Seattle, Washington.
- Llofriu, M., Tejera, G., Contreras, M., Fellous, J., & Weitzenfeld, A. (2015). Multi-scale Representation for Continuous State Q-Learning in Robotics. International Conference on Robotics and Automation, ICRA 2015. Seattle WA.
- Llofriu, M., Tejera, G., Contreras, M., Fellous, J., & Weitzenfeld, A. (2015, 11). Multi-scale Representation for Continuous State Q-Learning in Robotics. International Conference on Robotics and Automation, ICRA. Seattle WA.
- Malerba, P., Nagl, S., Krishnan, G., Fellous, J., & Bazhenov, M. (2015, 11). Understanding the mechanisms of hippocampal reactivation: from CA3 to CA1. Society for Neuroscience.
- Nagl, S., Crown, L., Jones, B., Tatsuno, M., & Fellous, J. (2015, 11). Sharp Wave Ripple Complexes Contribute to Context-Dependent Separation of Memories in a Rodent Reconsolidation Task.. Society for Neuroscience.
- Contreras, M., Llofriu, M., Weitzenfeld, A., & Fellous, J. (2014). Effect of dorsal or ventral hippocampus inactivation on goal-directed spatial navigation in rats and computational models. Society for Neuroscience. Washington DC.
- Janezic, E., Nagl, S., Uppalapati, S., French, E., & Fellous, J. (2014). Oxytocin and Social Bonding as Treatments in a Rodent Model of Post-Traumatic Stress Disorder. Society for Neuroscience. Washington DC.
- Lines, J., Nation, K., & Fellous, J. (2014). A connectionist model of context-based memory reconsolidation in the hippocampus: the role of sleep. Computational Neuroscience Meeting. Quebec City, Canada.
- Malerba, P., Krishnan, G. P., Fellous, J., & Bazhenov, M. (2014). Hippocampal ripples as inhibitory transients. Society for Neuroscience. Washington DC.