Orthogonal coding of object location

Activity in motor-sensory projections reveals distributed coding in somatosensation

Cortical feed-back projections to primary sensory areas terminate most heavily in layer (L) 1^1,2, where they make synapses with tuft dendrites of pyramidal neurons. L1 input is thought to provide ‘contextual’ information³, but the signals transmitted by L1 feedback remain uncharacterized. In the rodent somatosensory system, the spatially diffuse⁴ vibrissal motor cortex (vM1)→ vibrissal somatosensory cortex (barrel cortex, vS1) feedback projection may allow whisker touch to be interpreted in the context of whisker position to compute object location^5,6. When mice palpate objects with their whiskers to localize object features^7,8, whisker touch excites vibrissal somatosensory cortex (barrel cortex, vS1)⁹ and later vibrissal motor cortex (vM1) in a somatotopic manner^10,11,12,13. Here we used axonal calcium imaging to track activity in vM1→ vS1 afferents in L1 of barrel cortex, while mice performed whisker-dependent object localization. Spatially intermingled individual axons represented whisker movements, touch, and other behavioral features. In a subpopulation of axons, activity depended on object location and persisted for seconds after touch. Neurons in the barrel cortex thus have information to integrate movements and touches of multiple whiskers over time, key components of object identification and navigation by active touch.

Fast feedback in active sensing: touch-induced changes to whisker-object interaction

Whisking mediated touch is an active sense whereby whisker movements are modulated by sensory input and behavioral context. Here we studied the effects of touching an object on whisking in head-fixed rats. Simultaneous movements of whiskers C1, C2, and D1 were tracked bilaterally and their movements compared. During free-air whisking, whisker protractions were typically characterized by a single acceleration-deceleration event, whisking amplitude and velocity were correlated, and whisk duration correlated with neither amplitude nor velocity. Upon contact with an object, a second acceleration-deceleration event occurred in about 25% of whisk cycles, involving both contacting (C2) and non-contacting (C1, D1) whiskers ipsilateral to the object. In these cases, the rostral whisker (C2) remained in contact with the object throughout the double-peak phase, which effectively prolonged the duration of C2 contact. These “touch-induced pumps” (TIPs) were detected, on average, 17.9 ms after contact. On a slower time scale, starting at the cycle following first touch, contralateral amplitude increased while ipsilateral amplitude decreased. Our results demonstrate that sensory-induced motor modulations occur at various timescales, and directly affect object palpation.

Reinforcement active learning in the vibrissae system: optimal object localization

Rats move their whiskers to acquire information about their environment. It has been observed that they palpate novel objects and objects they are required to localize in space. We analyze whisker-based object localization using two complementary paradigms, namely, active learning and intrinsic-reward reinforcement learning. Active learning algorithms select the next training samples according to the hypothesized solution in order to better discriminate between correct and incorrect labels. Intrinsic-reward reinforcement learning uses prediction errors as the reward to an actor-critic design, such that behavior converges to the one that optimizes the learning process. We show that in the context of object localization, the two paradigms result in palpation whisking as their respective optimal solution. These results suggest that rats may employ principles of active learning and/or intrinsic reward in tactile exploration and can guide future research to seek the underlying neuronal mechanisms that implement them. Furthermore, these paradigms are easily transferable to biomimetic whisker-based artificial sensors and can improve the active exploration of their environment.

Dorsorostral snout muscles in the rat subserve coordinated movement for whisking and sniffing

Histochemical examination of the dorsorostral quadrant of the rat snout revealed superficial and deep muscles that are involved in whisking, sniffing, and airflow control. The part of M. nasolabialis profundus that acts as an intrinsic (follicular) muscle to facilitate protraction and translation of the vibrissae is described. An intraturbinate and selected rostral-most nasal muscles that can influence major routs of inspiratory airflow and rhinarial touch through their control of nostril configuration, atrioturbinate and rhinarium position, were revealed.

Motor-sensory convergence in object localization: a comparative study in rats and humans

In order to identify basic aspects in the process of tactile perception, we trained rats and humans in similar object localization tasks and compared the strategies used by the two species. We found that rats integrated temporally related sensory inputs ('temporal inputs') from early whisk cycles with spatially related inputs ('spatial inputs') to align their whiskers with the objects; their perceptual reports appeared to be based primarily on this spatial alignment. In a similar manner, human subjects also integrated temporal and spatial inputs, but relied mainly on temporal inputs for object localization. These results suggest that during tactile object localization, an iterative motor-sensory process gradually converges on a stable percept of object location in both species.

Multiple dynamic representations in the motor cortex during sensorimotor learning

The mechanisms linking sensation and action during learning are poorly understood. Layer 2/3 neurons in the motor cortex might participate in sensorimotor integration and learning; they receive input from sensory cortex and excite deep layer neurons, which control movement. Here we imaged activity in the same set of layer 2/3 neurons in the motor cortex over weeks, while mice learned to detect objects with their whiskers and report detection with licking. Spatially intermingled neurons represented sensory (touch) and motor behaviours (whisker movements and licking). With learning, the population-level representation of task-related licking strengthened. In trained mice, population-level representations were redundant and stable, despite dynamism of single-neuron representations. The activity of a subpopulation of neurons was consistent with touch driving licking behaviour. Our results suggest that ensembles of motor cortex neurons couple sensory input to multiple, related motor programs during learning.

Hierarchical curiosity loops and active sensing

A curious agent acts so as to optimize its learning about itself and its environment, without external supervision. We present a model of hierarchical curiosity loops for such an autonomous active learning agent, whereby each loop selects the optimal action that maximizes the agent's learning of sensory-motor correlations. The model is based on rewarding the learner's prediction errors in an actor-critic reinforcement learning (RL) paradigm. Hierarchy is achieved by utilizing previously learned motor-sensory mapping, which enables the learning of other mappings, thus increasing the extent and diversity of knowledge and skills. We demonstrate the relevance of this architecture to active sensing using the well-studied vibrissae (whiskers) system, where rodents acquire sensory information by virtue of repeated whisker movements. We show that hierarchical curiosity loops starting from optimally learning the internal models of whisker motion and then extending to object localization result in free-air whisking and object palpation, respectively.

Behavioral study of whisker-mediated vibration sensation in rats

Rats use their vibrissal sensory system to collect information about the nearby environment. They can accurately and rapidly identify object location, shape, and surface texture. Which features of whisker motion does the sensory system extract to construct sensations? We addressed this question by training rats to make discriminations between sinusoidal vibrations simultaneously presented to the left and right whiskers. One set of rats learned to reliably identify which of two vibrations had higher frequency (f(1) vs. f(2)) when amplitudes were equal. Another set of rats learned to reliably identify which of two vibrations had higher amplitude (A(1) vs. A(2)) when frequencies were equal. Although these results indicate that both elemental features contribute to the rats' sensation, a further test found that the capacity to discriminate A and f was reduced to chance when the difference in one feature was counterbalanced by the difference in the other feature: Rats could not discriminate amplitude or frequency whenever A(1)f(1) = A(2)f(2). Thus, vibrations were sensed as the product Af rather than as separable elemental features, A and f. The product Af is proportional to a physical entity, the mean speed. Analysis of performance revealed that rats extracted more information about differences in Af than predicted by the sum of the information in elemental differences. These behavioral experiments support the predictions of earlier physiological studies by demonstrating that rats are "blind" to the elemental features present in a sinusoidal whisker vibration; instead, they perceive a composite feature, the speed of whisker motion.

Neuronal basis for object location in the vibrissa scanning sensorimotor system

An essential issue in perception is how the location of an object is estimated from tactile signals in the context of self-generated changes in sensor configuration. Here, we review the pathways and dynamics of neuronal signals that encode touch in the rodent vibrissa sensorimotor system. Rodents rhythmically scan an array of long, facial hairs across a region of interest. Behavioral evidence shows that these animals maintain knowledge of the azimuthal position of their vibrissae. Electrophysiological measurements have identified a reafferent signal of the azimuth that is coded in normalized coordinates, broadcast throughout primary sensory cortex and provides strong modulation of signals of vibrissa contact. Efferent signals in motor cortex report the range of the scan. Collectively, these signals allow the rodent to form a percept of object location.

Psychometric curve and behavioral strategies for whisker-based texture discrimination in rats

The rodent whisker system is a major model for understanding neural mechanisms for tactile sensation of surface texture (roughness). Rats discriminate surface texture using the whiskers, and several theories exist for how texture information is physically sensed by the long, moveable macrovibrissae and encoded in spiking of neurons in somatosensory cortex. However, evaluating these theories requires a psychometric curve for texture discrimination, which is lacking. Here we trained rats to discriminate rough vs. fine sandpapers and grooved vs. smooth surfaces. Rats intermixed trials at macrovibrissa contact distance (nose >2 mm from surface) with trials at shorter distance (nose <2 mm from surface). Macrovibrissae were required for distant contact trials, while microvibrissae and non-whisker tactile cues were used for short distance trials. A psychometric curve was measured for macrovibrissa-based sandpaper texture discrimination. Rats discriminated rough P150 from smoother P180, P280, and P400 sandpaper (100, 82, 52, and 35 µm mean grit size, respectively). Use of olfactory, visual, and auditory cues was ruled out. This is the highest reported resolution for rodent texture discrimination, and constrains models of neural coding of texture information.

Chemosensory burst coding by mouse vomeronasal sensory neurons

The capabilities of any sensory system are ultimately constrained by the properties of the sensory neurons: the ability to detect and represent stimuli is limited by noise due to spontaneous activity, and optimal decoding in downstream circuitry must be matched to the nature of the encoding performed at the input. Here, we investigated the firing properties of sensory neurons in the accessory olfactory system, a distinct sensory system specialized for detection of socially relevant odors. Using multielectrode array recording, we observed that sensory neurons are spontaneously active and highly variable across time and trials and that this spontaneous activity limits the ability to distinguish sensory responses from noise. Sensory neuron activity tended to consist of bursts that maintained remarkably consistent statistics during both spontaneous activity and in response to stimulation with sulfated steroids. This, combined with pharmacological and genetic intervention in the signal transduction cascade, indicates that sensory transduction plays a role in shaping overall spontaneous activity. These findings indicate that as-yet unexplored characteristics of the sensory transduction cascade significantly constrain the representation of sensory information by vomeronasal neurons.

Toward an integrated approach to perception and action: conference report and future directions

This article was motivated by the conference entitled “Perception & Action – An Interdisciplinary Approach to Cognitive Systems Theory,” which took place September 14–16, 2010 at the Santa Fe Institute, NM, USA. The goal of the conference was to bring together an interdisciplinary group of neuroscientists, roboticists, and theorists to discuss the extent and implications of action–perception integration in the brain. The motivation for the conference was the realization that it is a widespread approach in biological, theoretical, and computational neuroscience to investigate sensory and motor function of the brain in isolation from one another, while at the same time, it is generally appreciated that sensory and motor processing cannot be fully separated. Our article summarizes the key findings of the conference, provides a hypothetical model that integrates the major themes and concepts presented at the conference, and concludes with a perspective on future challenges in the field.

Precision rodent whisker stimulator with integrated servo-locked control and displacement measurement

We developed a high speed voice coil based whisker stimulator that delivers precise deflections of a single whisker or group of whiskers in a repeatable manner. The device is miniature, quiet, and inexpensive to build. Multiple stimulators fit together for independent stimulation of four or more whiskers. The system can be used with animals under anesthesia as well as awake animals with head-restraint, and does not require trimming the whiskers. The system can deliver 1-2 mm deflections in 2 ms resulting in velocities up to 900 mm/s to attain a wide range of evoked responses. Since auditory artifacts can influence behavioral studies using whisker stimulation, we tested potential effects of auditory noise by recording somatosensory evoked potentials (SEP) with varying auditory click levels, and with/without 80 dBa background white noise. We found that auditory clicks as low as 40 dBa significantly influence the SEP. With background white noise, auditory clicks as low as 50 dBa were still detected in components of the SEP. For behavioral studies where animals must learn to respond to whisker stimulation, these sounds must be minimized. Together, the stimulator and data system can be used for psychometric vigilance tasks, mapping of the barrel cortex and other electrophysiological paradigms.

Role of the trigeminal mesencephalic nucleus in rat whisker pad proprioception

Background

Trigeminal proprioception related to rodent macrovibrissae movements is believed to involve skin receptors on the whisker pad because pad muscles operate without muscle spindles. This study was aimed to investigate in rats whether the trigeminal mesencephalic nucleus (TMnu), which provides proprioceptive feedback for chewing muscles, may be also involved in whisker pad proprioception.

Methods

Two retrograde tracers, Dil and True Blue Chloride, were injected into the mystacial pad and the masseter muscle on the same side of deeply anesthetized rats to label the respective projecting sensory neurons. This double-labeling technique was used to assess the co-innervation of both structures by the trigeminal mesencephalic nucleus (TMnu).

In a separate group of anesthetized animals, the spontaneous electrical activities of TMnu neurons were analyzed by extracellular recordings during spontaneous movements of the macrovibrissae. Mesencephalic neurons (TMne) were previously identified by their responses to masseter muscle stretching. Changes in TMne spontaneous electrical activities, analyzed under baseline conditions and during whisking movements, were statistically evaluated using Student's t-test for paired observations.

Results

Neuroanatomical experiments revealed different subpopulations of trigeminal mesencephalic neurons: i) those innervating the neuromuscular spindles of the masseter muscle, ii) those innervating the mystacial pad, and iii) those innervating both structures. Extracellular recordings made during spontaneous movements of the macrovibrisae showed that whisking neurons similar to those observed in the trigeminal ganglion were located in the TMnu. These neurons had different patterns of activation, which were dependent on the type of spontaneous macrovibrissae movement. In particular, their spiking activity tonically increased during fan-like movements of the vibrissae and showed phasic bursting during rhythmic whisking. Furthermore, the same neurons may also respond to masseter muscle stretch.

Conclusions

results strongly support the hypothesis that the TMnu also contains first-order neurons specialized for relaying spatial information related to whisker movement and location to trigeminal-cortical pathways. In fact, the TMnu projects to second-order trigeminal neurons, thus allowing the rat brain to deduce higher-order information regarding executed movements of the vibrissae by combining touch information carried by trigeminal ganglion neurons with proprioceptive information carried by mesencephalic neurons.

A comparison of neuronal and behavioral detection and discrimination performances in rat whisker system

We used the rat whisker touch as a model system to investigate the correlation between the response function of cortical neurons and the behavior of rats in a sensory detection versus discrimination task. The rat whisker-barrel system is structurally well characterized and represents one of the main channels through which rodents collect information about the environment. In experiment 1, we recorded neuronal activity (n = 235) in the whisker area of the rat somatosensory cortex in anesthetized rats while applying vibrotactile stimuli of varying amplitudes to the whiskers. Neurons showed a characteristic sigmoidal input-output function, with an accelerating nonlinearity at low stimulus amplitudes and a compressive nonlinearity at high stimulus amplitudes. We further quantified the performance of individual neurons for stimulus detection and for discrimination across different stimulus pairs with identical amplitude differences. For near-threshold stimuli, the neuronal discrimination performance surpassed the detection performance despite the fact that detection and discrimination represented identical amplitude differences. This is consistent with the accelerating nonlinearity observed at low stimulus intensities. In the second stage of the experiment, four rats were trained to select the higher-amplitude stimulus between two vibrations applied to their whiskers. Similar to neuronal results, the rats' performance was better for the discrimination task compared with the detection task. The behavioral performance followed the same trend as that of the population of individual neurons. Both behavioral and neuronal data are consistent with the "pedestal effect" previously reported in human psychophysics.

Spiking activity propagation in neuronal networks: reconciling different perspectives on neural coding

The brain is a highly modular structure. To exploit modularity, it is necessary that spiking activity can propagate from one module to another while preserving the information it carries. Therefore, reliable propagation is one of the key properties of a candidate neural code. Surprisingly, the conditions under which spiking activity can be propagated have received comparatively little attention in the experimental literature. By contrast, several computational studies in the last decade have addressed this issue. Using feedforward networks (FFNs) as a generic network model, they have identified two dynamical activity modes that support the propagation of either asynchronous (rate code) or synchronous (temporal code) spiking. Here, we review the dichotomy of asynchronous and synchronous propagation in FFNs, propose their integration into a single extended conceptual framework and suggest experimental strategies to test our hypothesis.

Rat psychomotor vigilance task with fast response times using a conditioned lick behavior

Investigations into the physiological mechanisms of sleep control require an animal psychomotor vigilance task (PVT) with fast response times (<300 ms). Rats provide a good PVT model since whisker stimulation produces a rapid and robust cortical evoked response, and animals can be trained to lick following stimulation. Our prior experiments used deprivation-based approaches to maximize motivation for operant conditioned responses. However, deprivation can influence physiological and neurobehavioral effects. In order to maintain motivation without water deprivation, we conditioned rats for immobilization and head restraint, then trained them to lick for a 10% sucrose solution in response to whisker stimulation. After approximately 8 training sessions, animals produced greater than 80% correct hits to the stimulus. Over the course of training, reaction times became faster and correct hits increased. Performance in the PVT was examined after 3, 6 and 12 h of sleep deprivation achieved by gentle handling. A significant decrease in percent correct hits occurred following 6 and 12 h of sleep deprivation and reaction times increased significantly following 12 h of sleep deprivation. While behaviorally the animals appeared to be awake, we observed significant increases in EEG delta power prior to misses. The rat PVT with fast response times allows investigation of sleep deprivation effects, time-on-task and pharmacological agents. Fast response times also allow closer parallel studies to ongoing human protocols.

Information Carried by Population Spike Times in the Whisker Sensory Cortex can be Decoded Without Knowledge of Stimulus Time

Computational analyses have revealed that precisely timed spikes emitted by somatosensory cortical neuronal populations encode basic stimulus features in the rat's whisker sensory system. Efficient spike time based decoding schemes both for the spatial location of a stimulus and for the kinetic features of complex whisker movements have been defined. To date, these decoding schemes have been based upon spike times referenced to an external temporal frame – the time of the stimulus itself. Such schemes are limited by the requirement of precise knowledge of the stimulus time signal, and it is not clear whether stimulus times are known to rats making sensory judgments. Here, we first review studies of the information obtained from spike timing referenced to the stimulus time. Then we explore new methods for extracting spike train information independently of any external temporal reference frame. These proposed methods are based on the detection of stimulus-dependent differences in the firing time within a neuronal population. We apply them to a data set using single-whisker stimulation in anesthetized rats and find that stimulus site can be decoded based on the millisecond-range relative differences in spike times even without knowledge of stimulus time. If spike counts alone are measured over tens or hundreds of milliseconds rather than milliseconds, such decoders are much less effective. These results suggest that decoding schemes based on millisecond-precise spike times are likely to subserve robust and information-rich transmission of information in the somatosensory system.

Between molecules and experience: role of early patterns of coordinated activity for the development of cortical maps and sensory abilities

Sensory systems processing information from the environment rely on precisely formed and refined neuronal networks that build maps of sensory receptor epithelia at different subcortical and cortical levels. These sensory maps share similar principles of function and emerge according to developmental processes common in visual, somatosensory and auditory systems. Whereas molecular cues set the coarse organization of cortico-subcortical topography, its refinement is known to succeed under the influence of experience-dependent electrical activity during critical periods. However, coordinated patterns of activity synchronize the cortico-subcortical networks long before the meaningful impact of environmental inputs on sensory maps. Recent studies elucidated the cellular and network mechanisms underlying the generation of these early patterns of activity and highlighted their similarities across species. Moreover, the experience-independent activity appears to act as a functional template for the maturation of sensory networks and cortico-subcortical maps. A major goal for future research will be to analyze how this early activity interacts with the molecular cues and to determine whether it is permissive or rather supporting for the establishment of sensory topography.

Texture coding in the whisker system

The whisker somatosensory system in awake, behaving rodents is a powerful model for studying neurobiology of sensation, from molecules to circuits to behavior. Recent studies reveal how key tactile features are detected in awake animals and encoded by spike trains in somatosensory cortex (S1). Here we summarize progress on detection of surface texture (roughness). Texture appears to be inferred from the statistics of complex, irregular whisker micromotion on surfaces, specifically by mean speed or by patterns of discrete, high-velocity whisker slips. These are encoded in S1 by mean firing rate and by sparse, synchronous, slip-evoked spike volleys, respectively. An alternative model of place coding for texture based on differential whisker resonance is less well supported, but is not ruled out.

Sensory neural codes using multiplexed temporal scales

Determining how neuronal activity represents sensory information is central for understanding perception. Recent work shows that neural responses at different timescales can encode different stimulus attributes, resulting in a temporal multiplexing of sensory information. Multiplexing increases the encoding capacity of neural responses, enables disambiguation of stimuli that cannot be discriminated at a single response timescale, and makes sensory representations stable to the presence of variability in the sensory world. Thus, as we discuss here, temporal multiplexing could be a key strategy used by the brain to form an information-rich and stable representation of the environment.

Stability of neural firing in the trigeminal nuclei under mechanical whisker stimulation

Sensory information handling is an essentially nonstationary process even under a periodic stimulation. We show how the time evolution of ridges in the wavelet spectrum of spike trains can be used for quantification of the dynamical stability of the neuronal responses to a stimulus. We employ this method to study neuronal responses in trigeminal nuclei of the rat provoked by tactile whisker stimulation. Neurons from principalis (Pr5) and interpolaris (Sp5i) show the maximal stability at the intermediate (50 ms) stimulus duration, whereas Sp5o cells "prefer" shorter (10 ms) stimulation. We also show that neurons in all three nuclei can perform as stimulus frequency filters. The response stability of about 33% of cells exhibits low-pass frequency dynamics. About 57% of cells have band-pass dynamics with the optimal frequency at 5 Hz for Pr5 and Sp5i, and 4 Hz for Sp5o, and the remaining 10% show no prominent dependence on the stimulus frequency. This suggests that the neural coding scheme in trigeminal nuclei is not fixed, but instead it adapts to the stimulus characteristics.