Goal-oriented robot navigation learning using a multi-scale space representation.

There has been extensive research in recent years on the multi-scale nature of hippocampal place cells and entorhinal grid cells encoding which led to many speculations on their role in spatial cognition. In this paper we focus on the multi-scale nature of place cells and how they contribute to faster learning during goal-oriented navigation when compared to a spatial cognition system composed of single scale place cells. The task consists of a circular arena with a fixed goal location, in which a robot is trained to find the shortest path to the goal after a number of learning trials. Synaptic connections are modified using a reinforcement learning paradigm adapted to the place cells multi-scale architecture. The model is evaluated in both simulation and physical robots. We find that larger scale and combined multi-scale representations favor goal-oriented navigation task learning.

Synaptic background noise controls the input/output characteristics of single cells in an in vitro model of in vivo activity.

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 alpha-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 gamma-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.

A new correlation-based measure of spike timing reliability.

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.

Attractor reliability reveals deterministic structure in neuronal spike trains.

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.

Information transfer in entrained cortical neurons.

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.

Fluctuating synaptic conductances recreate in vivo-like activity in neocortical neurons.

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.

Frequency dependence of spike timing reliability in cortical pyramidal cells and interneurons.

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. Cortical 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.

The brain decade in debate: III. Neurobiology of emotion.

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.

The brain decade in debate: III. Neurobiology of emotion

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

Computational model of carbachol-induced delta, theta, and gamma oscillations in the hippocampus.

Field potential recordings from the rat hippocampus in vivo contain distinct frequency bands of activity, including delta (0.5-2 Hz), theta (4-12 Hz), and gamma (30-80 Hz), that are correlated with the behavioral state of the animal. The cholinergic agonist carbachol (CCH) induces oscillations in the delta (CCH-delta), theta (CCH-theta), and gamma (CCH-gamma) frequency ranges in the hippocampal slice preparation, eliciting asynchronous CCH-theta, synchronous CCH-delta, and synchronous CCH-theta with increasing CCH concentration (Fellous and Seinowski, Hippocampus 2000;1 0:187-197). In a network model of area CA3, the time scale for CCH-delta corresponded to the decay constant of the gating variable of the calcium-dependent potassium (K-AHP) current, that of CCH-theta to an intrinsic subthreshold membrane potential oscillation of the pyramidal cells, and that of CCH-gamma 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-theta to CCH-delta and from CCH-delta to synchronous CCH-theta. The simulations also exhibited the interspersed CCH-gamma/CCH-delta and CCH-gamma/CCH-theta 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.

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.

Carbachol, a muscarinic receptor agonist, produced three distinct spontaneous oscillations in the CA3 region of rat hippocampal slices. Carbachol concentrations in the 4-13 microM range produced regular synchronized CA3 discharges at 0.5-2 Hz (carbachol-delta). Higher concentrations (13-60 microM) produced short episodes of 5-10 Hz (carbachol-theta) oscillations separated by nonsynchronous activity. Concentrations of carbachol ranging from 8-25 microM 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.

Computational models of neuromodulation.

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.

Gender discrimination and prediction on the basis of facial metric information.

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.