Bulletin of the American Physical Society
2009 APS March Meeting
Volume 54, Number 1
Monday–Friday, March 16–20, 2009; Pittsburgh, Pennsylvania
Session Q40: Neural Computation |
Hide Abstracts |
Sponsoring Units: DBP Chair: John Beggs, Indiana University Room: 412 |
Wednesday, March 18, 2009 11:15AM - 11:27AM |
Q40.00001: Deriving functional structure of neuronal networks from spike train data Sarah Feldt, Vaughn Hetrick, Joshua Berke, Michal Zochowski We present a novel algorithm for the detection of functional clusters in neural data. In contrast to many clustering techniques which convert functional interactions to topological distances to determine groupings, our algorithm directly utilizes the dynamics of the neurons to obtain functional groupings. No prior knowledge of the number of groups is needed, as the algorithm determines statistically significant clusters through a comparison to surrogate data sets. Additionally, we introduce a new synchronization measure and use this measure in the algorithm to observe known groupings in simulated data. We then apply our algorithm to experimental data obtained from the hippocampus of a freely moving mouse and show that it detects known changes in neural states associated with exploration and slow wave sleep. Finally, we show that the new synchronization measure can detect changes which are consistent with known neurophysiological processes involved in memory consolidation. [Preview Abstract] |
Wednesday, March 18, 2009 11:27AM - 11:39AM |
Q40.00002: Bursting dynamics of in vitro neural networks and their stimulation driven learning. Joon Ho Choi, June Hoan Kim, Kyoung J. Lee Recent studies have indicated that recurring neural ``bursts'' may play an essential role in neural information processing and memory. One key element of this hypothesis involves the translation of temporal patterns of stimuli into spatiotemporally distributed information. One ideal system to investigate this issue is cultured network of neurons grown on multi-electrode array (MEA). Based on such in vitro systems, we have investigated the changes incurred by extrinsic stimuli in the spontaneously recurring bursting activities. We have employed, in particular, two-channel paired, delayed, tetanic stimuli to evoke different patterns of bursting activities. Our preliminary data suggests that the neural network can exhibit some learning behavior. [Preview Abstract] |
Wednesday, March 18, 2009 11:39AM - 11:51AM |
Q40.00003: Burst switching between incoherence and synchrony Nathan Crosby, Joseph Tranquillo Studies of coupled oscillators often use diffusive connections to ensure that the coupled quantity is conserved. Signals between neurons, however, are not diffusive and may propagate unattenuated throughout a network. We compare diffusive and synapse-like coupling of Hindmarsh-Rose (HMR) oscillators through numerical simulations. HMR parameters are tuned to either oscillate continuously or alternate bursts of oscillations and periods of quiescence. In diffusive coupling, two HMR units synchronize bursts and individual oscillations within a burst at nearly the same coupling strength. Synapse-like coupling, however, shows a new behavior, called burst switching, between incoherence and synchrony. For example, a bursting unit can entrain an oscillating unit of a different frequency during the burst but then force the oscillator into quiescence. Burst switching in various network topologies synchronizes inhomogeneous units for the duration of the burst, followed by a period of network quiescence and a return to incoherence. The summed activity resembles the progression of an epileptic seizure including the ``spike and wave'' at the transition from synchrony to quiescence. [Preview Abstract] |
Wednesday, March 18, 2009 11:51AM - 12:03PM |
Q40.00004: ABSTRACT WITHDRAWN |
Wednesday, March 18, 2009 12:03PM - 12:15PM |
Q40.00005: Trajectories through similarity space produced by local neocortical circuits John Beggs, Wei Chen, Jon Hobbs, Aonan Tang The dynamics found in local cortical networks strongly impact the types of computations they can perform. Major classes of cortical network models assume that spatio-temporal activity evolves with either ultra-stable, chaotic or neutral dynamics. While experimental evidence has demonstrated that repeatable activity states can exist in cortical networks, it is still unclear what the spatio-temporal dynamics near these states are. To accurately address this question, the trajectories of similar, but not identical, inputs must be quantified. We use 60 channel microelectrode arrays to measure spatio-temporal trajectories through similarity space at 4 ms resolution in organotypic cortical cultures and acute cortical slices. Here we show that while attractive, chaotic and neutral trajectories can exist in these networks, the average trajectory has a Lyapunov exponent near zero (0.01 $\pm $ 0.2, mean $\pm $ s.d.), indicating that neutral dynamics prevail. [Preview Abstract] |
Wednesday, March 18, 2009 12:15PM - 12:27PM |
Q40.00006: Synaptic weight distribution under spike-timing dependent plasticity Chun-Chung Chen, David Jasnow We consider a network of integrate-and-fire neurons with random connections driven by noise triggered firings. The synaptic weights between the neurons are allowed to evolve under spike-timing dependent plasticity rules with additive potentiation and multiplicative depression. Under realistic physiological parameters, the network was equilibrated with simulations for a physical time of days. For lower potentiation-to-depression ratio $w^{\star}$, the synaptic weights forms a unimodal distribution which decays for large weights following a power law with a strong negative exponent. The decay exponent increases with $w^{\star}$, and runaway synaptic weights were observed as the exponent approaches $-1$. In the stationary state under the plasticity, for low $w^{\star}$, triggering the firing of a single neuron in a quiet network typically leads to a bursting event that lasts for seconds in a small network of 32 neurons. For high $w^{\star}$, the induced activities can persist in the network indefinitely. A mean-field theory combined with a master equation describing the distribution of synaptic weights predicts a power-law regime under the small jump assumption of synaptic weight changes. The exponents of predicted power law depends on the deviation of the mean synaptic weight from the $w^{\star}$ parameter and is to be determined self-consistently. [Preview Abstract] |
Wednesday, March 18, 2009 12:27PM - 12:39PM |
Q40.00007: Hebbian based learning with winner-take-all for spiking neural networks Ankur Gupta, Lyle Long Learning methods for spiking neural networks are not as well developed as the traditional neural networks that widely use back-propagation training. We propose and implement a Hebbian based learning method with winner-take-all competition for spiking neural networks. This approach is spike time dependent which makes it naturally well suited for a network of spiking neurons. Homeostasis with Hebbian learning is implemented which ensures stability and quicker learning. Homeostasis implies that the net sum of incoming weights associated with a neuron remains the same. Winner-take-all is also implemented for competitive learning between output neurons. We implemented this learning rule on a biologically based vision processing system that we are developing, and use layers of leaky integrate and fire neurons. The network when presented with 4 bars (or Gabor filters) of different orientation learns to recognize the bar orientations (or Gabor filters). After training, each output neuron learns to recognize a bar at specific orientation and responds by firing more vigorously to that bar and less vigorously to others. These neurons are found to have bell shaped tuning curves and are similar to the simple cells experimentally observed by Hubel and Wiesel in the striate cortex of cat and monkey. [Preview Abstract] |
Wednesday, March 18, 2009 12:39PM - 12:51PM |
Q40.00008: Neural Decision Boundaries Predict Maximum Entropy Parameters Jeff Fitzgerald, Tatyana Sharpee Previous studies have shown that the response properties of neural networks can be well described by a pairwise maximum entropy model (PMEM). The coupling constants in this model can be calculated from experimental data, but it is unknown how they would need to change to optimally encode different distributions of stimuli. To determine the optimal coupling constants for a given stimulus distribution, we extended the model of neural decision boundaries to networks of neurons. This model of neural responses assumes stimuli that elicit a spike are separated from those that do not by a decision boundary. We demonstrate that the coupling constants of the PMEM which maximize information can be found from smoothness conditions on the decision boundaries. We considered exponentially distributed stimuli that mimic the large deviations found in signals in the natural environment and found that the optimal coupling constants between pairs of neurons are indeed non-zero, in agreement with experimental data. [Preview Abstract] |
Wednesday, March 18, 2009 12:51PM - 1:03PM |
Q40.00009: Relationship between higher-order correlations in stimulus and information in the receptive fields of visual neurons Ryan Rowekamp, Tatyana Sharpee Neurons encode incoming signals in a series of spikes in the voltage trace across their cell membranes. This encoding is known to change in response to stimulus mean, variance, and power spectrum. Natural signals are known to have strong higher-than-second order correlations that cannot be described by a Gaussian distribution. To examine whether these higher-order statistics can also cause neurons to adapt their codes, we modeled the neural spike probability as an arbitrary nonlinear function with respect to two stimulus dimensions. The relevant stimulus dimensions were found as those that accounted for the largest mutual information between stimuli and spikes. We found that the contribution of the second dimension on the spike probability was stronger for natural, rather than Gaussian noise, stimuli and increased with the kurtosis of the stimulus distribution. [Preview Abstract] |
Wednesday, March 18, 2009 1:03PM - 1:15PM |
Q40.00010: Robust Motion Processing in the Visual Cortex Audrey Sederberg, Julia Liu, Matthias Kaschube Direction selectivity is an important model system for studying cortical processing. The role of inhibition in models of direction selectivity in the visual cortex is not well understood. We probe the selectivity of an integrate-and-fire neuron with a noisy background on top of a deterministic input current determined by a temporal-lag model for selectivity, including first only excitatory inputs and later both excitatory and inhibitory input. In this model, postsynaptic potentials are fully synchronous for the preferred direction and maximally dispersed in time for the null direction. Further, any inhibitory inputs lag excitatory inputs, as Priebe and Ferster have observed (2005). At any level of input strength, the selectivity is weak when only excitatory inputs are considered. The inclusion of inhibition significantly strengthens selectivity, and this selectivity is preserved over a wide range of background noise levels and for short stimulus durations. We conclude that inhibition likely plays an essential role in the mechanism underlying direction selectivity. [Preview Abstract] |
Wednesday, March 18, 2009 1:15PM - 1:27PM |
Q40.00011: Rhythmogenic Neuronal Networks and k-Core Percolation David Schwab, Robijn Bruinsma, Alex Levine The \textit{preB\"{o}tzinger Complex} (pBC) is a small ($\sim $10$^{2})$ network of identical excitatory neurons that collectively generate a temporally stable pattern of firing bursts interspersed by quiescent periods. The voltage output of this system is essential to the control of the mammalian breathing rhythm under certain physiological conditions. The network is also remarkable in that a small set of coupled identical neurons can generate a collective behavior that is not inherent in any one of them: individual neurons do not exhibit rhythmic bursting. We develop a simple model of interacting excitatory neurons that demonstrates this behavior as one of its dynamical regimes, and show that while some of its dynamical transitions can be understood in terms of mean field theory, others cannot. The non-mean-field behavior can be understood in terms of purely topological properties of random networks. [Preview Abstract] |
Wednesday, March 18, 2009 1:27PM - 1:39PM |
Q40.00012: ABSTRACT WITHDRAWN |
Wednesday, March 18, 2009 1:39PM - 1:51PM |
Q40.00013: Determining information flow in networks containing one hundred neocortical neurons Aonan Tang, Jon Hobbs, Wladek Dabrowski, Pawel Hottowy, Alexander Sher, Alan Litke, John Beggs How does information flow through networks of neurons? The type of network topology revealed could have important consequences for network efficiency and robustness to damage. Several tools, including transfer entropy, Granger causality, and directed information can be applied to this question. Yet indirect connections, connections with various delays, and feedback loops can complicate the task of uncovering the information flow structure. We have applied the above methods in simple validation studies, demonstrating that many of these issues can in principle be overcome. We will present preliminary results from neocortical networks of neurons recorded with a 512 electrode array. [Preview Abstract] |
Wednesday, March 18, 2009 1:51PM - 2:03PM |
Q40.00014: Integration of neuroblasts into a two-dimensional small world neuronal network Casey Schneider-Mizell, Michal Zochowski, Leonard Sander Neurogenesis in the adult brain has been suggested to be important for learning and functional robustness to the neuronal death. New neurons integrate themselves into existing neuronal networks by moving into a target destination, extending axonal and dendritic processes, and inducing synaptogenesis to connect to active neurons. We hypothesize that increased plasticity of the network to novel stimuli can arise from activity-dependent cell and process motility rules. In complement to a similar in vitro model, we investigate a computational model of a two-dimensional small world network of integrate and fire neurons. After steady-state activity is reached in the extant network, we introduce new neurons which move, stop, and connect themselves through rules governed by position and firing rate. [Preview Abstract] |
Wednesday, March 18, 2009 2:03PM - 2:15PM |
Q40.00015: Phase dependent and independent responses in auditory cortex Didier Depireux, Barak Shechter Responses of auditory neurons are often characterized by their spectro-temporal receptive field (STRF). This linear measure has been shown to capture the overall trend of the response, but by its nature, it does not reflect any nonlinear processing. We have recently shown that neurons in primary auditory cortex (AI) of the awake ferret respond with non-trivial nonlinearities (not solely the result of rectifying or saturating nonlinearities). We developed new techniques to reveal additional phase independent (DC) and dependent (quadratic) tuning in the tuning of single neurons. One of the assumptions in the STRF model is that the mean firing rate (averaged over any single period of the stimulus) does not depend on the spectro-temporal modulations, but rather on the overall level of the stimulus. The phase-independent tuning to the spectro-temporal envelope is analogous to complex visual neural responses, in which responses to an auditory grating stimulus do not depend on its spatial phase. We show the existence of neurons tuned in 1) a phase-independent manner, 2) a linear manner and 3) a quadratic manner to the time-frequency content of the spectral envelope of sounds. [Preview Abstract] |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2025 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700