Flutter discrimination: neural codes, perception, memory and decision making

Sensory-motor interactions modulate a primate vocal behavior: antiphonal calling in common marmosets

A fundamental issue in neuroscience pertains to how different cortical systems interact to generate behavior. One of the most direct ways to address this issue is to investigate how sensory information is encoded and used to produce a motor response. Antiphonal calling is a natural vocal behavior that involves individuals producing their species-specific long distance vocalization in response to hearing the same call and engages both the auditory and motor systems, as well as the cognitive neural systems involved in decision making and categorization. Here we present results from a series of behavioral experiments investigating the auditory-vocal interactions during antiphonal calling in the common marmoset (Callithrix jacchus). We manipulated sensory input by placing subjects in different social contexts and found that the auditory input had a significant effect on call timing and propensity to call. Playback experiments tested the significance of the timing of vocal production in antiphonal calling and showed that a short latency between antiphonal calls was necessary to maintain reciprocal vocal interactions. Overall, this study shows that sensory-motor interactions can be experimentally induced and manipulated in a natural primate vocal behavior. Antiphonal calling represents a promising model system to examine these issues in non-human primates at both the behavioral and neural levels.

Neural codes for perceptual discrimination in primary somatosensory cortex

We sought to determine the neural code(s) for frequency discrimination of vibrotactile stimuli. We tested five possible candidate codes by analyzing the responses of single neurons recorded in primary somatosensory cortex of trained monkeys while they discriminated between two consecutive vibrotactile stimuli. Differences in the frequency of two stimuli could be discriminated using information from (i) time intervals between spikes, (ii) average spiking rate during each stimulus, (iii) absolute number of spikes elicited by each stimulus, (iv) average rate of bursts of spikes or (v) absolute number of spike bursts elicited by each stimulus. However, only a spike count code, in which spikes are integrated over a time window that has most of its mass in the first 250 ms of each stimulus period, covaried with behavior on a trial-by-trial basis, was consistent with psychophysical biases induced by manipulation of stimulus duration, and produced neurometric discrimination thresholds similar to behavioral psychophysical thresholds.

Approaches to the study of haptic sensing

This review surveys results from a new approach to the problem of haptic sensing, in which subjects use primarily proximal arm movements to explore the shapes of virtual objects. These shapes are generated using a robotically controlled manipulandum. We begin by summarizing distortions of simple geometric properties (such as the length and orientation of lines) in the haptic perception of space. We then consider the extent to which the sense of more complex shapes (such as quadrilaterals) can be explained by these geometric distortions, i.e., the extent to which the shape of a complex object is synthesized from simpler constituent elements, and some of the sensory cues that may be important in this process. Haptic and visual processing of shapes appear to lead to some similar illusions. However, we argue that the processing of haptic information differs fundamentally from visual processing in that the former requires the integration of information that evolves in time as well as in space.

Patterned firing of parietal cells in a haptic working memory task

Abstract Cells in the somatosensory cortex of the monkey are known to exhibit sustained elevations of firing frequency during the short-term mnemonic retention of tactile information in a haptic delay task. In this study, we examine the possibility that those firing elevations are accompanied by changes in firing pattern. Patterns are identified by the application of a pattern-searching algorithm to the interspike intervals of spike trains. By sequential use of sets of pattern templates with a range of temporal resolutions, we find patterned activity in the majority of the cells investigated. In general, the degree of patterning significantly increases during active memory. Surrogate analysis suggests that the observed patterns may not be simple linear stochastic functions of instantaneous or average firing frequency. Therefore, during the active retention of a memorandum, the activity of a 'memory cell' may be characterized not only by changes in frequency but also by changes in pattern.

Learning of somatosensory representations for texture discrimination using a temporal coherence principle

In order to perform appropriate actions, animals need to quickly and reliably classify their sensory input. How can representations suitable for classification be acquired from statistical properties of the animal's natural environment? Akin to behavioural studies in rats, we investigate this question using texture discrimination by the vibrissae system as a model. To account for the rat's active sensing behaviour, we record whisker movements in a hardware model. Based on these signals, we determine the response of primary neurons, modelled as spatio-temporal filters. Using their output, we train a second layer of neurons to optimise a temporal coherence objective function. The performance in classifying textures using a single cell strongly correlates with the cell's temporal coherence; hence output cells outperform primary cells. Using a simple, unsupervised classifier, the performance on the output cell population is same as if using a sophisticated supervised classifier on the primary cells. Our results demonstrate that the optimisation of temporal coherence yields a representation that facilitates subsequent classification by selectively conveying relevant information.

Choosing the greater of two goods: neural currencies for valuation and decision making

To make adaptive decisions, animals must evaluate the costs and benefits of available options. The nascent field of neuroeconomics has set itself the ambitious goal of understanding the brain mechanisms that are responsible for these evaluative processes. A series of recent neurophysiological studies in monkeys has begun to address this challenge using novel methods to manipulate and measure an animal's internal valuation of competing alternatives. By emphasizing the behavioural mechanisms and neural signals that mediate decision making under conditions of uncertainty, these studies might lay the foundation for an emerging neurobiology of choice behaviour.

Quantitative prediction of perceptual decisions during near-threshold fear detection

A fundamental goal of cognitive neuroscience is to explain how mental decisions originate from basic neural mechanisms. The goal of the present study was to investigate the neural correlates of perceptual decisions in the context of emotional perception. To probe this question, we investigated how fluctuations in functional MRI (fMRI) signals were correlated with behavioral choice during a near-threshold fear detection task. fMRI signals predicted behavioral choice independently of stimulus properties and task accuracy in a network of brain regions linked to emotional processing: posterior cingulate cortex, medial prefrontal cortex, right inferior frontal gyrus, and left insula. We quantified the link between fMRI signals and behavioral choice in a whole-brain analysis by determining choice probabilities by means of signal-detection theory methods. Our results demonstrate that voxel-wise fMRI signals can reliably predict behavioral choice in a quantitative fashion (choice probabilities ranged from 0.63 to 0.78) at levels comparable to neuronal data. We suggest that the conscious decision that a fearful face has been seen is represented across a network of interconnected brain regions that prepare the organism to appropriately handle emotionally challenging stimuli and that regulate the associated emotional response.

Tactile estimation of the roughness of gratings yields a graded response in the human brain: an fMRI study

Human subjects can tactually estimate the magnitude of surface roughness. Although many psychophysical and neurophysiological experiments have elucidated the peripheral neural mechanisms that underlie tactile roughness estimation, the associated cortical mechanisms are not well understood. To identify the brain regions responsible for the tactile estimation of surface roughness, we used functional magnetic resonance imaging (fMRI). We utilized a combination of categorical (subtraction) and parametric factorial approaches wherein roughness was varied during both the task and its control. Fourteen human subjects performed a tactile roughness-estimation task and received the identical tactile stimulation without estimation (no-estimation task). The bilateral parietal operculum (PO), insula and right lateral prefrontal cortex showed roughness-related activation. The bilateral PO and insula showed activation during the no-estimation task, and hence might represent the sensory-based processing during roughness estimation. By contrast, the right prefrontal cortex is more related to the cognitive processing, as there was activation during the estimation task compared with the no-estimation task, but little activation was observed during the no-estimation task in comparison with rest. The lateral prefrontal area might play an important cognitive role in tactile estimation of surface roughness, whereas the PO and insula might be involved in the sensory processing that is important for estimating surface roughness.

Reinforcement learning and decision making in monkeys during a competitive game

Animals living in a dynamic environment must adjust their decision-making strategies through experience. To gain insights into the neural basis of such adaptive decision-making processes, we trained monkeys to play a competitive game against a computer in an oculomotor free-choice task. The animal selected one of two visual targets in each trial and was rewarded only when it selected the same target as the computer opponent. To determine how the animal's decision-making strategy can be affected by the opponent's strategy, the computer opponent was programmed with three different algorithms that exploited different aspects of the animal's choice and reward history. When the computer selected its targets randomly with equal probabilities, animals selected one of the targets more often, violating the prediction of probability matching, and their choices were systematically influenced by the choice history of the two players. When the computer exploited only the animal's choice history but not its reward history, animal's choice became more independent of its own choice history but was still related to the choice history of the opponent. This bias was substantially reduced, but not completely eliminated, when the computer used the choice history of both players in making its predictions. These biases were consistent with the predictions of reinforcement learning, suggesting that the animals sought optimal decision-making strategies using reinforcement learning algorithms.

Flexible Control of Mutual Inhibition: A Neural Model of Two-Interval Discrimination

Networks adapt to environmental demands by switching between distinct dynamical behaviors. The activity of frontal-lobe neurons during two-interval discrimination tasks is an example of these adaptable dynamics. Subjects first perceive a stimulus, then hold it in working memory, and finally make a decision by comparing it with a second stimulus. We present a simple mutual-inhibition network model that captures all three task phases within a single framework. The model integrates both working memory and decision making because its dynamical properties are easily controlled without changing its connectivity. Mutual inhibition between nonlinear units is a useful design motif for networks that must display multiple behaviors.

Working memory in primate sensory systems

Sensory working memory consists of the short-term storage of sensory stimuli to guide behaviour. There is increasing evidence that elemental sensory dimensions - such as object motion in the visual system or the frequency of a sound in the auditory system - are stored by segregated feature-selective systems that include not only the prefrontal and parietal cortex, but also areas of sensory cortex that carry out relatively early stages of processing. These circuits seem to have a dual function: precise sensory encoding and short-term storage of this information. New results provide insights into how activity in these circuits represents the remembered sensory stimuli.

Lentivirus-based genetic manipulations of cortical neurons and their optical and electrophysiological monitoring in vivo

It is becoming increasingly clear that single cortical neurons encode complex and behaviorally relevant signals, but efficient means to study gene functions in small networks and single neurons in vivo are still lacking. Here, we establish a method for genetic manipulation and subsequent phenotypic analysis of individual cortical neurons in vivo. First, lentiviral vectors are used for neuron-specific gene delivery from alpha-calcium/calmodulin-dependent protein kinase II or Synapsin I promoters, optionally in combination with gene knockdown by means of U6 promoter-driven expression of short-interfering RNAs. Second, the phenotypic analysis at the level of single cortical cells is carried out by using two-photon microscopy-based techniques: high-resolution two-photon time-lapse imaging is used to monitor structural dynamics of dendritic spines and axonal projections, whereas cellular response properties are analyzed electrophysiologically by two-photon microscopy directed whole-cell recordings. This approach is ideally suited for analysis of gene functions in individual neurons in the intact brain.

Representation of Number in Animals and Humans: A Neural Model

This article addresses the representation of numerical information conveyed by nonsymbolic and symbolic stimuli. In a first simulation study, we show how number-selective neurons develop when an initially uncommitted neural network is given nonsymbolic stimuli as input (e.g., collections of dots) under unsupervised learning. The resultant network is able to account for the distance and size effects, two ubiquitous effects in numerical cognition. Furthermore, the properties of the network units conform in detail to the characteristics of recently discovered number-selective neurons. In a second study, we simulate symbol learning by presenting symbolic and nonsymbolic input simultaneously. The same number-selective neurons learn to represent the numerical meaning of symbols. In doing so, they show properties reminiscent of the originally available number-selective neurons, but at the same time, the representational efficiency of the neurons is increased when presented with symbolic input. This finding presents a concrete proposal on the linkage between higher order numerical cognition and more primitive numerical abilities and generates specific predictions on the neural substrate of number processing.

A general mechanism for perceptual decision-making in the human brain

Findings from single-cell recording studies suggest that a comparison of the outputs of different pools of selectively tuned lower-level sensory neurons may be a general mechanism by which higher-level brain regions compute perceptual decisions. For example, when monkeys must decide whether a noisy field of dots is moving upward or downward, a decision can be formed by computing the difference in responses between lower-level neurons sensitive to upward motion and those sensitive to downward motion. Here we use functional magnetic resonance imaging and a categorization task in which subjects decide whether an image presented is a face or a house to test whether a similar mechanism is also at work for more complex decisions in the human brain and, if so, where in the brain this computation might be performed. Activity within the left dorsolateral prefrontal cortex is greater during easy decisions than during difficult decisions, covaries with the difference signal between face- and house-selective regions in the ventral temporal cortex, and predicts behavioural performance in the categorization task. These findings show that even for complex object categories, the comparison of the outputs of different pools of selectively tuned neurons could be a general mechanism by which the human brain computes perceptual decisions.

Task-specific recruitment of dorsal and ventral visual areas during tactile perception

Many studies have found that visual cortical areas are active during tactile perception. Here we used positron emission tomographic (PET) scanning in normally sighted humans to show that extrastriate cortical regions are recruited in a task-specific manner during perceptual processing of tactile stimuli varying in two dimensions. Mental rotation of tactile Forms activated a focus around the anterior part of the left intraparietal sulcus. Since prior studies have reported activity nearby during mental rotation of visual stimuli, this focus appears to be associated with the dorsal visual (visuospatial) pathway. Discrimination between tactile Forms activated the right lateral occipital complex, an object-selective region in the ventral visual (visual Form) pathway. Thus, tactile tasks appear to recruit cortical regions that are active during corresponding visual tasks. Activation of these areas in both visual and tactile tasks could reflect visual imagery during tactile perception, activity in multisensory representations, or both.

On building a bridge between brain and behavior

Cognitive neuroscience is motivated by the precept that a discoverable correspondence exists between mental states and brain states. This precept seems to be supported by remarkable observations and conclusions derived from event-related potentials and functional imaging with humans and neurophysiology with behaving monkeys. This review evaluates specific conceptual and technical limits of claims of correspondence between neural events, overt behavior, and hypothesized covert processes examined using data on the neural control of saccadic eye movements.

Neuronal Correlates of a Perceptual Decision in Ventral Premotor Cortex

The ventral premotor cortex (VPC) is involved in the transformation of sensory information into action, although the exact neuronal operation is not known. We addressed this problem by recording from single neurons in VPC while trained monkeys report a decision based on the comparison of two mechanical vibrations applied sequentially to the fingertips. Here we report that the activity of VPC neurons reflects current and remembered sensory inputs, their comparison, and motor commands expressing the result; that is, the entire processing cascade linking the evaluation of sensory stimuli with a motor report. These findings provide a fairly complete panorama of the neural dynamics that underlies the transformation of sensory information into an action and emphasize the role of VPC in perceptual decisions.

Neuronal activity in primate dorsolateral and orbital prefrontal cortex during performance of a reward preference task

An important function of the prefrontal cortex (PFC) is the control of goal-directed behaviour. This requires information as to whether actions were successful in obtaining desired outcomes such as rewards. While lesion studies implicate a particular PFC region, the orbitofrontal cortex (OFC), in reward processing, neurons encoding reward have been reported in both the OFC and the dorsolateral prefrontal cortex (DLPFC). To compare and contrast their roles, we recorded simultaneously from both areas while two rhesus monkeys (Macaca mulatta) performed a reward preference task. The monkeys had to choose between pictures associated with different amounts of a juice reward. Neuronal activity in both areas reflected the reward amount. However, neurons in the DLPFC encoded both the reward amount and the monkeys' forthcoming response, while neurons in the OFC more often encoded the reward amount alone. Further, reward selectivity arose more rapidly in the OFC than the DLPFC. These results are consistent with reward information entering the PFC via the OFC, where it is passed to the DLPFC and used to control behaviour.

Haptic synthesis of shapes and sequences

Haptic perception of shape is based on kinesthetic and tactile information synthesized across space and time. We studied this process by having subjects move along the edges of multisided shapes and then remember and reproduce the shapes. With eyes closed, subjects moved a robot manipulandum whose force field was programmed to simulate a quadrilateral boundary in a horizontal plane. When subjects then reproduced the quadrilateral using the same manipulandum, with eyes still closed but now with the force field set to zero, they made consistent errors, overestimating the lengths of short segments and underestimating long ones, as well as overestimating acute angles and underestimating obtuse ones. Consequently their reproductions were more regular than the shapes they had experienced. When subjects felt the same quadrilaterals with the same manipulandum but drew them on a vertical screen with visual feedback, they made similar errors, indicating that their distortions reflected mainly perceptual rather than motor processes. In a third experiment, subjects explored the 3 sides of an open shape in a fixed order. The results revealed a temporal pattern of interactions, where the lengths and angles of previously explored segments influenced the drawing of later segments. In all tasks, our subjects were as accurate as subjects in earlier studies who haptically explored only single lines or angles, suggesting that the mental processes that synthesize haptic data from multiple segments into complete shapes do not introduce any net error.

A comparison of primate prefrontal and inferior temporal cortices during visual categorization

Previous studies have suggested that both the prefrontal cortex (PFC) and inferior temporal cortex (ITC) are involved in high-level visual processing and categorization, but their respective roles are not known. To address this, we trained monkeys to categorize a continuous set of visual stimuli into two categories, "cats" and "dogs." The stimuli were parametrically generated using a computer graphics morphing system (Sheltonelton, 2000) that allowed precise control over stimulus shape. After training, we recorded neural activity from the PFC and the ITC of monkeys while they performed a category-matching task. We found that the PFC and the ITC play distinct roles in category-based behaviors: the ITC seems more involved in the analysis of currently viewed shapes, whereas the PFC showed stronger category signals, memory effects, and a greater tendency to encode information in terms of its behavioral meaning.

Microstimulation of visual cortex affects the speed of perceptual decisions

Direction-selective neurons in the middle temporal visual area (MT) are crucially involved in motion perception, although it is not known exactly how the activity of these neurons is interpreted by the rest of the brain. Here we report that in a two-alternative task, the activity of MT neurons is interpreted as evidence for one direction and against the other. We measured the speed and accuracy of decisions as rhesus monkeys performed a direction-discrimination task. On half of the trials, we stimulated direction-selective neurons in area MT, thereby causing the monkeys to choose the neurons' preferred direction more often. Microstimulation quickened decisions in favor of the preferred direction and slowed decisions in favor of the opposite direction. Even on trials in which microstimulation did not induce a preferred direction choice, it still affected response times. Our findings suggest that during the formation of a decision, sensory evidence for competing propositions is compared and accumulates to a decision-making threshold.