What is all the noise about in interval timing?

Cognitive processes such as decision-making, rate calculation and planning require an accurate estimation of durations in the supra-second range—interval timing. In addition to being accurate, interval timing is scale invariant: the time-estimation errors are proportional to the estimated duration. The origin and mechanisms of this fundamental property are unknown. We discuss the computational properties of a circuit consisting of a large number of (input) neural oscillators projecting on a small number of (output) coincidence detector neurons, which allows time to be coded by the pattern of coincidental activation of its inputs. We showed analytically and checked numerically that time-scale invariance emerges from the neural noise. In particular, we found that errors or noise during storing or retrieving information regarding the memorized criterion time produce symmetric, Gaussian-like output whose width increases linearly with the criterion time. In contrast, frequency variability produces an asymmetric, long-tailed Gaussian-like output, that also obeys scale invariant property. In this architecture, time-scale invariance depends neither on the details of the input population, nor on the distribution probability of noise.     

Minimal models of multidimensional computations

The multidimensional computations performed by many biological systems are often characterized with limited information about the correlations between inputs and outputs. Given this limitation, our approach is to construct the maximum noise entropy response function of the system, leading to a closed-form and minimally biased model consistent with a given set of constraints on the input/output moments; the result is equivalent to conditional random field models from machine learning. For systems with binary outputs, such as neurons encoding sensory stimuli, the maximum noise entropy models are logistic functions whose arguments depend on the constraints. A constraint on the average output turns the binary maximum noise entropy models into minimum mutual information models, allowing for the calculation of the information content of the constraints and an information theoretic characterization of the system's computations. We use this approach to analyze the nonlinear input/output functions in macaque retina and thalamus; although these systems have been previously shown to be responsive to two input dimensions, the functional form of the response function in this reduced space had not been unambiguously identified. A second order model based on the logistic function is found to be both necessary and sufficient to accurately describe the neural responses to naturalistic stimuli, accounting for an average of 93% of the mutual information with a small number of parameters. Thus, despite the fact that the stimulus is highly non-Gaussian, the vast majority of the information in the neural responses is related to first and second order correlations. Our results suggest a principled and unbiased way to model multidimensional computations and determine the statistics of the inputs that are being encoded in the outputs.     

Optimal feedback control to describe multiple representations of primary motor cortex neurons

Primary motor cortex (M1) neurons are tuned in response to several parameters related to motor control, and it was recently reported that M1 is important in feedback control. However, it remains unclear how M1 neurons encode information to control the musculoskeletal system. In this study, we examined the underlying computational mechanisms of M1 based on optimal feedback control (OFC) theory, which is a plausible hypothesis for neuromotor control. We modelled an isometric torque production task that required joint torque to be regulated and maintained at desired levels in a musculoskeletal system physically constrained by muscles, which act by pulling rather than pushing. Then, a feedback controller was computed using an optimisation approach under the constraint. In the presence of neuromotor noise, known as signal-dependent noise, the sensory feedback gain is tuned to an extrinsic motor output, such as the hand force, like a population response of M1 neurons. Moreover, a distribution of the preferred directions (PDs) of M1 neurons can be predicted via feedback gain. Therefore, we suggest that neural activity in M1 is optimised for the musculoskeletal system. Furthermore, if the feedback controller is represented in M1, OFC can describe multiple representations of M1, including not only the distribution of PDs but also the response of the neuronal population.     

Role of uncertainty in sensorimotor control

Neural signals are corrupted by noise and this places limits on information processing. We review the processes involved in goal-directed movements and how neural noise and uncertainty determine aspects of our behaviour. First, noise in sensory signals limits perception. We show that, when localizing our hand, the central nervous system (CNS) integrates visual and proprioceptive information, each with different noise properties, in a way that minimizes the uncertainty in the overall estimate. Second, noise in motor commands leads to inaccurate movements. We review an optimal-control framework, known as 'task optimization in the presence of signal-dependent noise', which assumes that movements are planned so as to minimize the deleterious consequences of noise and thereby minimize inaccuracy. Third, during movement, sensory and motor signals have to be integrated to allow estimation of the body's state. Models are presented that show how these signals are optimally combined. Finally, we review how the CNS deals with noise at the neural and network levels. In all of these processes, the CNS carries out the tasks in such a way that the detrimental effects of noise are minimized. This shows that it is important to consider effects at the neural level in order to understand performance at the behavioural level.     

Evaluation and amelioration of computer-aided diagnosis with artificial neural networks utilizing small-sized sample sets

Concerns about the specificity and reliability of artificial neural networks (ANNs) impede further application of ANNs in medicine. This is particularly true when developing computer-aided diagnosis (CAD) tools using ANNs for orphan diseases and emerging research areas where only a small-sized sample set is available. It is unreasonable to claim one ANN's performance as better than another simply on the basis of a single output without considering possible output variability due to factors including data noise and ANN training protocols. In this paper, a bootstrap resampling method is proposed to quantitatively analyze ANN output reliability and changing performance as the sample data and training protocols are varied. The method is tested in the area of feature classification for analysis of masses detected on mammograms. Our experiments show that ANNs performance, measured in terms of the area under the receiver operating characteristic (ROC) curve, is not a fixed value, but follows a distribution function sensitive to many factors. We demonstrate that our approach to determining the bootstrap estimates of confidence intervals (CIs) and prediction intervals (PIs) can be used to assure optimal performance in terms of ANN model configuration. We also show that the unintentional inclusion of data noise, which biases ANN results in small task-specific databases, can be accurately detected via the bootstrap estimates.     

A generative model for measuring latent timing structure in motor sequences

Motor variability often reflects a mixture of different neural and peripheral sources operating over a range of timescales. We present a statistical model of sequence timing that can be used to measure three distinct components of timing variability: global tempo changes that are spread across the sequence, such as might stem from neuromodulatory sources with widespread influence; fast, uncorrelated timing noise, stemming from noisy components within the neural system; and timing jitter that does not alter the timing of subsequent elements, such as might be caused by variation in the motor periphery or by measurement error. In addition to quantifying the variability contributed by each of these latent factors in the data, the approach assigns maximum likelihood estimates of each factor on a trial-to-trial basis. We applied the model to adult zebra finch song, a temporally complex behavior with rich structure on multiple timescales. We find that individual song vocalizations (syllables) contain roughly equal amounts of variability in each of the three components while overall song length is dominated by global tempo changes. Across our sample of syllables, both global and independent variability scale with average length while timing jitter does not, a pattern consistent with the Wing and Kristofferson (1973) model of sequence timing. We also find significant day-to-day drift in all three timing sources, but a circadian pattern in tempo only. In tests using artificially generated data, the model successfully separates out the different components with small error. The approach provides a general framework for extracting distinct sources of timing variability within action sequences, and can be applied to neural and behavioral data from a wide array of systems.     

A neuromorphic architecture for object recognition and motion anticipation using burst-STDP

In this work we investigate the possibilities offered by a minimal framework of artificial spiking neurons to be deployed in silico. Here we introduce a hierarchical network architecture of spiking neurons which learns to recognize moving objects in a visual environment and determine the correct motor output for each object. These tasks are learned through both supervised and unsupervised spike timing dependent plasticity (STDP). STDP is responsible for the strengthening (or weakening) of synapses in relation to pre- and post-synaptic spike times and has been described as a Hebbian paradigm taking place both in vitro and in vivo. We utilize a variation of STDP learning, called burst-STDP, which is based on the notion that, since spikes are expensive in terms of energy consumption, then strong bursting activity carries more information than single (sparse) spikes. Furthermore, this learning algorithm takes advantage of homeostatic renormalization, which has been hypothesized to promote memory consolidation during NREM sleep. Using this learning rule, we design a spiking neural network architecture capable of object recognition, motion detection, attention towards important objects, and motor control outputs. We demonstrate the abilities of our design in a simple environment with distractor objects, multiple objects moving concurrently, and in the presence of noise. Most importantly, we show how this neural network is capable of performing these tasks using a simple leaky-integrate-and-fire (LIF) neuron model with binary synapses, making it fully compatible with state-of-the-art digital neuromorphic hardware designs. As such, the building blocks and learning rules presented in this paper appear promising for scalable fully neuromorphic systems to be implemented in hardware chips.     

Timing as an intrinsic property of neural networks: evidence from in vivo and in vitro experiments

The discrimination and production of temporal patterns on the scale of hundreds of milliseconds are critical to sensory and motor processing. Indeed, most complex behaviours, such as speech comprehension and production, would be impossible in the absence of sophisticated timing mechanisms. Despite the importance of timing to human learning and cognition, little is known about the underlying mechanisms, in particular whether timing relies on specialized dedicated circuits and mechanisms or on general and intrinsic properties of neurons and neural circuits. Here, we review experimental data describing timing and interval-selective neurons in vivo and in vitro. We also review theoretical models of timing, focusing primarily on the state-dependent network model, which proposes that timing in the subsecond range relies on the inherent time-dependent properties of neurons and the active neural dynamics within recurrent circuits. Within this framework, time is naturally encoded in populations of neurons whose pattern of activity is dynamically changing in time. Together, we argue that current experimental and theoretical studies provide sufficient evidence to conclude that at least some forms of temporal processing reflect intrinsic computations based on local neural network dynamics.     

The sniff as a unit of olfactory processing

Sniffing is a rhythmic motor process essential for the acquisition of olfactory information. Recent behavioral experiments show that using a single sniff rats can accurately discriminate between very similar odors and fail to improve their accuracy by taking multiple sniffs. This implies that each sniff has the potential to provide a complete snapshot of the local olfactory environment. The discrete and intermittent nature of sniffing has implications beyond the physical process of odor capture as it strongly shapes the flow of information into the olfactory system. We review electrophysiological studies-primarily from anesthetized rodents-demonstrating that olfactory neural responses are coupled to respiration. Hence, the "sniff cycle" might play a role in odor coding, by allowing the timing of spikes with respect to the phase of the respiration cycle to encode information about odor identity or concentration. We also discuss behavioral and physiological results indicating that sniffing can be dynamically coordinated with other rhythmic behaviors, such as whisking, as well as with rhythmic neural activity, such as hippocampal theta oscillations. Thus, the sniff cycle might also facilitate the coordination of the olfactory system with other brain areas. These converging lines of empirical data support the notion that each sniff is a unit of olfactory processing relevant for both neural coding and inter-areal coordination. Further electrophysiological recordings in behaving animals will be necessary to assess these proposals.     

Evolving coordinated group behaviours through maximisation of mean mutual information

A well known problem in the design of the control system for a swarm of robots concerns the definition of suitable individual rules that result in the desired coordinated behaviour. A possible solution to this problem is given by the automatic synthesis of the individual controllers through evolutionary or learning processes. These processes offer the possibility to freely search the space of the possible solutions for a given task, under the guidance of a user-defined utility function. Nonetheless, there exist no general principles to follow in the definition of such a utility function in order to reward coordinated group behaviours. As a consequence, task dependent functions must be devised each time a new coordination problem is under study. In this paper, we propose the use of measures developed in Information Theory as task-independent, implicit utility functions. We present two experiments in which three robots are trained to produce generic coordinated behaviours. Each robot is provided with rich sensory and motor apparatus, which can be exploited to explore the environment and to communicate with other robots. We show how coordinated behaviours can be synthesised through a simple evolutionary process. The only criteria used to evaluate the performance of the robotic group is the estimate of mutual information between the motor states of the robots.     

Order in spontaneous behavior

Brains are usually described as input/output systems: they transform sensory input into motor output. However, the motor output of brains (behavior) is notoriously variable, even under identical sensory conditions. The question of whether this behavioral variability merely reflects residual deviations due to extrinsic random noise in such otherwise deterministic systems or an intrinsic, adaptive indeterminacy trait is central for the basic understanding of brain function. Instead of random noise, we find a fractal order (resembling Lévy flights) in the temporal structure of spontaneous flight maneuvers in tethered Drosophila fruit flies. Lévy-like probabilistic behavior patterns are evolutionarily conserved, suggesting a general neural mechanism underlying spontaneous behavior. Drosophila can produce these patterns endogenously, without any external cues. The fly's behavior is controlled by brain circuits which operate as a nonlinear system with unstable dynamics far from equilibrium. These findings suggest that both general models of brain function and autonomous agents ought to include biologically relevant nonlinear, endogenous behavior-initiating mechanisms if they strive to realistically simulate biological brains or out-compete other agents.     

New moves in motor control

Motor behaviour results from information processing across multiple neural networks acting at all levels from initial selection of the behaviour to its final generation. Understanding how motor behaviour is produced requires identifying the constituent neurons of these networks, their cellular properties, and their pattern of synaptic connectivity. Neural networks have been traditionally studied with neurophysiological and neuroanatomical approaches. These approaches have been highly successful in particularly suitable 'model' preparations, typically ones in which the numbers of neurons in the networks were relatively small, neural network composition was unvarying across individual animals, and the preparations continued to produce fictive motor patterns in?vitro. However, analysing networks without these characteristics, and analysing the complete ensemble of networks that cooperatively generate behaviours, is difficult with these approaches. Recently developed molecular and neurogenetic tools provide additional avenues for analysing motor networks by allowing individual or groups of neurons within networks to be manipulated in novel ways and allowing experiments to be performed not only in?vitro but also in?vivo. We review here some of the new insights into motor network function that these advances have provided and indicate how these advances might bridge gaps in our understanding of motor control. To these ends, we first review motor neural network organisation highlighting cross-phylum principles. We then use prominent examples from the field to show how neurogenetic approaches can complement classical physiological studies, and identify additional areas where these approaches could be advantageously applied.     

The Main Sequence of Saccades Optimizes Speed-accuracy Trade-off

In primates, it is well known that there is a consistent relationship between the duration, peak velocity and amplitude of saccadic eye movements, known as the ‘main sequence’. The reason why such a stereotyped relationship evolved is unknown. We propose that a fundamental constraint on the deployment of foveal vision lies in the motor system that is perturbed by signal-dependent noise (proportional noise) on the motor command. This noise imposes a compromise between the speed and accuracy of an eye movement. We propose that saccade trajectories have evolved to optimize a trade-off between the accuracy and duration of the movement. Taking a semi-analytical approach we use Pontryagin’s minimum principle to show that there is an optimal trajectory for a given amplitude and duration; and that there is an optimal duration for a given amplitude. It follows that the peak velocity is also fixed for a given amplitude. These predictions are in good agreement with observed saccade trajectories and the main sequence. Moreover, this model predicts a small saccadic dead-zone in which it is better to stay eccentric of target than make a saccade onto target. We conclude that the main sequence has evolved as a strategy to optimize the trade-off between accuracy and speed.     

Variability and coding efficiency of noisy neural spike encoders

Encoding synaptic inputs as a train of action potentials is a fundamental function of nerve cells. Although spike trains recorded in vivo have been shown to be highly variable, it is unclear whether variability in spike timing represents faithful encoding of temporally varying synaptic inputs or noise inherent in the spike encoding mechanism. It has been reported that spike timing variability is more pronounced for constant, unvarying inputs than for inputs with rich temporal structure. This could have significant implications for the nature of neural coding, particularly if precise timing of spikes and temporal synchrony between neurons is used to represent information in the nervous system. To study the potential functional role of spike timing variability, we estimate the fraction of spike timing variability which conveys information about the input for two types of noisy spike encoders--an integrate and fire model with randomly chosen thresholds and a model of a patch of neuronal membrane containing stochastic Na(+) and K(+) channels obeying Hodgkin-Huxley kinetics. The quality of signal encoding is assessed by reconstructing the input stimuli from the output spike trains using optimal linear mean square estimation. A comparison of the estimation performance of noisy neuronal models of spike generation enables us to assess the impact of neuronal noise on the efficacy of neural coding. The results for both models suggest that spike timing variability reduces the ability of spike trains to encode rapid time-varying stimuli. Moreover, contrary to expectations based on earlier studies, we find that the noisy spike encoding models encode slowly varying stimuli more effectively than rapidly varying ones.     

An approximate stochastic optimal control framework to simulate nonlinear neuro-musculoskeletal models in the presence of noise

Optimal control simulations have shown that both musculoskeletal dynamics and physiological noise are important determinants of movement. However, due to the limited efficiency of available computational tools, deterministic simulations of movement focus on accurately modelling the musculoskeletal system while neglecting physiological noise, and stochastic simulations account for noise while simplifying the dynamics. We took advantage of recent approaches where stochastic optimal control problems are approximated using deterministic optimal control problems, which can be solved efficiently using direct collocation. We were thus able to extend predictions of stochastic optimal control as a theory of motor coordination to include muscle coordination and movement patterns emerging from non-linear musculoskeletal dynamics. In stochastic optimal control simulations of human standing balance, we demonstrated that the inclusion of muscle dynamics can predict muscle co-contraction as minimal effort strategy that complements sensorimotor feedback control in the presence of sensory noise. In simulations of reaching, we demonstrated that nonlinear multi-segment musculoskeletal dynamics enables complex perturbed and unperturbed reach trajectories under a variety of task conditions to be predicted. In both behaviors, we demonstrated how interactions between task constraint, sensory noise, and the intrinsic properties of muscle influence optimal muscle coordination patterns, including muscle co-contraction, and the resulting movement trajectories. Our approach enables a true minimum effort solution to be identified as task constraints, such as movement accuracy, can be explicitly imposed, rather than being approximated using penalty terms in the cost function. Our approximate stochastic optimal control framework predicts complex features, not captured by previous simulation approaches, providing a generalizable and valuable tool to study how musculoskeletal dynamics and physiological noise may alter neural control of movement in both healthy and pathological movements.     

On the definition of central pattern generator and its sensory control

A central pattern generator (CPG) is defined here as a neural network responsible for the production of the timing cues of a rhythmic motor output pattern in the isolated CNS. For the intact animal, model considerations show that this term is neither clearly delimited from the concept of a reflex chain nor from the concept of a pattern generator with functional principles different from those of the CPG. Therefore, it cannot be concluded from the existence of a CPG in the isolated nervous system that this CPG also provides the decisive timing cues in the intact animal. Consequences for the study of the neural basis of rhythmic movements are shown.     

Millisecond-Scale Motor Encoding in a Cortical Vocal Area

Studies of motor control have almost universally examined firing rates to investigate how the brain shapes behavior. In principle, however, neurons could encode information through the precise temporal patterning of their spike trains as well as (or instead of) through their firing rates. Although the importance of spike timing has been demonstrated in sensory systems, it is largely unknown whether timing differences in motor areas could affect behavior. We tested the hypothesis that significant information about trial-by-trial variations in behavior is represented by spike timing in the songbird vocal motor system. We found that neurons in motor cortex convey information via spike timing far more often than via spike rate and that the amount of information conveyed at the millisecond timescale greatly exceeds the information available from spike counts. These results demonstrate that information can be represented by spike timing in motor circuits and suggest that timing variations evoke differences in behavior.     

Neural mechanisms underlying the evolvability of behaviour

The complexity of nervous systems alters the evolvability of behaviour. Complex nervous systems are phylogenetically constrained; nevertheless particular species-specific behaviours have repeatedly evolved, suggesting a predisposition towards those behaviours. Independently evolved behaviours in animals that share a common neural architecture are generally produced by homologous neural structures, homologous neural pathways and even in the case of some invertebrates, homologous identified neurons. Such parallel evolution has been documented in the chromatic sensitivity of visual systems, motor behaviours and complex social behaviours such as pair-bonding. The appearance of homoplasious behaviours produced by homologous neural substrates suggests that there might be features of these nervous systems that favoured the repeated evolution of particular behaviours. Neuromodulation may be one such feature because it allows anatomically defined neural circuitry to be re-purposed. The developmental, genetic and physiological mechanisms that contribute to nervous system complexity may also bias the evolution of behaviour, thereby affecting the evolvability of species-specific behaviour.     

Implications of conspecific background noise for features of blue tit, Cyanistes caeruleus , communication networks at dawn

Communication among animals often comprises several signallers and receivers within the signal’s transmission range. In such communication networks, individuals can extract information about differences in relative performance of conspecifics by eavesdropping on their signalling interactions. In songbirds, information can be encoded in the timing of signals, which either alternate or overlap, and both male and female receivers may utilise this information when engaging in territorial interactions or making reproductive decisions, respectively. We investigated how conspecific background noise at dawn may overlay and possibly constrain the perception of such singing patterns. We simulated a small communication network of blue tits, Cyanistes caeruleus, at dawn in spring. Two loudspeakers simulated a singing interaction which was recorded from four different receiver positions simulating potential eavesdroppers. During the recordings, resident blue tit males were vocalising and created natural conspecific background noise. Levels of conspecific background noise were high and varied among positions of potential eavesdroppers. We conclude that conspecific background vocalisations may potentially constrain the perception of singing patterns and may constitute costs for eavesdroppers. On the other hand, signallers may position themselves strategically and privatise their interactions.     

A case for spiking neural network simulation based on configurable multiple-FPGA systems

Recent neuropsychological research has begun to reveal that neurons encode information in the timing of spikes. Spiking neural network simulations are a flexible and powerful method for investigating the behaviour of neuronal systems. Simulation of the spiking neural networks in software is unable to rapidly generate output spikes in large-scale of neural network. An alternative approach, hardware implementation of such system, provides the possibility to generate independent spikes precisely and simultaneously output spike waves in real time, under the premise that spiking neural network can take full advantage of hardware inherent parallelism. We introduce a configurable FPGA-oriented hardware platform for spiking neural network simulation in this work. We aim to use this platform to combine the speed of dedicated hardware with the programmability of software so that it might allow neuroscientists to put together sophisticated computation experiments of their own model. A feed-forward hierarchy network is developed as a case study to describe the operation of biological neural systems (such as orientation selectivity of visual cortex) and computational models of such systems. This model demonstrates how a feed-forward neural network constructs the circuitry required for orientation selectivity and provides platform for reaching a deeper understanding of the primate visual system. In the future, larger scale models based on this framework can be used to replicate the actual architecture in visual cortex, leading to more detailed predictions and insights into visual perception phenomenon.