Inferring control objectives in a virtual balancing task in humans and monkeys.

Natural behaviors have redundancy, which implies that humans and animals can achieve their goals with different strategies. Given only observations of behavior, is it possible to infer the control objective that the subject is employing? This challenge is particularly acute in animal behavior because we cannot ask or instruct the subject to use a particular strategy. This study presents a three-pronged approach to infer an animal's control objective from behavior. First, both humans and monkeys performed a virtual balancing task for which different control strategies could be utilized. Under matched experimental conditions, corresponding behaviors were observed in humans and monkeys. Second, a generative model was developed that represented two main control objectives to achieve the task goal. Model simulations were used to identify aspects of behavior that could distinguish which control objective was being used. Third, these behavioral signatures allowed us to infer the control objective used by human subjects who had been instructed to use one control objective or the other. Based on this validation, we could then infer objectives from animal subjects. Being able to positively identify a subject's control objective from observed behavior can provide a powerful tool to neurophysiologists as they seek the neural mechanisms of sensorimotor coordination.

An ethologically motivated neurobiology of primate visually-guided reach-to-grasp behavior.

The precision of primate visually guided reaching likely evolved to meet the many challenges faced by living in arboreal environments, yet much of what we know about the underlying primate brain organization derives from a set of highly constrained experimental paradigms. Here we review the role of vision to guide natural reach-to-grasp movements in marmoset monkey prey capture to illustrate the breadth and diversity of these behaviors in ethological contexts, the fast predictive nature of these movements [1,2], and the advantages of this particular primate model to investigate the underlying neural mechanisms in more naturalistic contexts [3]. In addition to their amenability to freely-moving neural recording methods for investigating the neural basis of dynamic ethological behaviors [4,5], marmosets have a smooth neocortical surface that facilitates imaging and array recordings [6,7] in all areas in the primate fronto-parietal network [8,9]. Together, this model organism offers novel opportunities to study the real-world interplay between primate vision and reach-to-grasp dynamics using ethologically motivated neuroscientific experimental designs.

Learning leaves a memory trace in motor cortex.

How are we able to learn new behaviors without disrupting previously learned ones? To understand how the brain achieves this, we used a brain-computer interface (BCI) learning paradigm, which enables us to detect the presence of a memory of one behavior while performing another. We found that learning to use a new BCI map altered the neural activity that monkeys produced when they returned to using a familiar BCI map in a way that was specific to the learning experience. That is, learning left a "memory trace" in the primary motor cortex. This memory trace coexisted with proficient performance under the familiar map, primarily by altering neural activity in dimensions that did not impact behavior. Forming memory traces might be how the brain is able to provide for the joint learning of multiple behaviors without interference.

Motor cortex retains and reorients neural dynamics during motor imagery.

The most prominent characteristic of motor cortex is its activation during movement execution, but it is also active when we simply imagine movements in the absence of actual motor output. Despite decades of behavioural and imaging studies, it is unknown how the specific activity patterns and temporal dynamics in motor cortex during covert motor imagery relate to those during motor execution. Here we recorded intracortical activity from the motor cortex of two people who retain some residual wrist function following incomplete spinal cord injury as they performed both actual and imagined isometric wrist extensions. We found that we could decompose the population activity into three orthogonal subspaces, where one was similarly active during both action and imagery, and the others were active only during a single task type-action or imagery. Although they inhabited orthogonal neural dimensions, the action-unique and imagery-unique subspaces contained a strikingly similar set of dynamic features. Our results suggest that during motor imagery, motor cortex maintains the same overall population dynamics as during execution by reorienting the components related to motor output and/or feedback into a unique, output-null imagery subspace.

Dynamical constraints on neural population activity.

The manner in which neural activity unfolds over time is thought to be central to sensory, motor, and cognitive functions in the brain. Network models have long posited that the brain's computations involve time courses of activity that are shaped by the underlying network. A prediction from this view is that the activity time courses should be difficult to violate. We leveraged a brain-computer interface (BCI) to challenge monkeys to violate the naturally-occurring time courses of neural population activity that we observed in motor cortex. This included challenging animals to traverse the natural time course of neural activity in a time-reversed manner. Animals were unable to violate the natural time courses of neural activity when directly challenged to do so. These results provide empirical support for the view that activity time courses observed in the brain indeed reflect the underlying network-level computational mechanisms that they are believed to implement.

Common Mechanisms of Learning in Motor and Cognitive Systems.

Rapid progress in our understanding of the brain's learning mechanisms has been accomplished over the past decade, particularly with conceptual advances, including representing behavior as a dynamical system, large-scale neural population recordings, and new methods of analysis of neuronal populations. However, motor and cognitive systems have been traditionally studied with different methods and paradigms. Recently, some common principles, evident in both behavior and neural activity, that underlie these different types of learning have become to emerge. Here we review results from motor and cognitive learning, relying on different techniques and studying different systems to understand the mechanisms of learning. Movement is intertwined with cognitive operations, and its dynamics reflect cognitive variables. Training, in either motor or cognitive tasks, involves recruitment of previously unresponsive neurons and reorganization of neural activity in a low dimensional manifold. Mapping of new variables in neural activity can be very rapid, instantiating flexible learning of new tasks. Communication between areas is just as critical a part of learning as are patterns of activity within an area emerging with learning. Common principles across systems provide a map for future research.

Inferring control objectives in a virtual balancing task in humans and monkeys.

Natural behaviors have redundancy, which implies that humans and animals can achieve their goals with different control objectives. Given only observations of behavior, is it possible to infer the control strategy that the subject is employing? This challenge is particularly acute in animal behavior because we cannot ask or instruct the subject to use a particular control strategy. This study presents a threepronged approach to infer an animal's control strategy from behavior. First, both humans and monkeys performed a virtual balancing task for which different control objectives could be utilized. Under matched experimental conditions, corresponding behaviors were observed in humans and monkeys. Second, a generative model was developed that represented two main control strategies to achieve the task goal. Model simulations were used to identify aspects of behavior that could distinguish which control objective was being used. Third, these behavioral signatures allowed us to infer the control objective used by human subjects who had been instructed to use one control objective or the other. Based on this validation, we could then infer strategies from animal subjects. Being able to positively identify a subject's control objective from behavior can provide a powerful tool to neurophysiologists as they seek the neural mechanisms of sensorimotor coordination.

A neural basis of choking under pressure.

Incentives tend to drive improvements in performance. But when incentives get too high, we can "choke under pressure" and underperform when it matters most. What neural processes might lead to choking under pressure? We studied Rhesus monkeys performing a challenging reaching task in which they underperform when an unusually large "jackpot" reward is at stake. We observed a collapse in neural information about upcoming movements for jackpot rewards: in the motor cortex, neural planning signals became less distinguishable for different reach directions when a jackpot reward was made available. We conclude that neural signals of reward and motor planning interact in the motor cortex in a manner that can explain why we choke under pressure. One-Sentence Summary: In response to exceptionally large reward cues, animals can "choke under pressure", and this corresponds to a collapse in the neural information about upcoming movements.

Motor cortex retains and reorients neural dynamics during motor imagery.

The most prominent role of motor cortex is generating patterns of neural activity that lead to movement, but it is also active when we simply imagine movements in the absence of actual motor output. Despite decades of behavioral and imaging studies, it is unknown how the specific activity patterns and temporal dynamics within motor cortex during covert motor imagery relate to those during motor execution. Here we recorded intracortical activity from the motor cortex of two people with residual wrist function following incomplete spinal cord injury as they performed both actual and imagined isometric wrist extensions. We found that we could decompose the population-level activity into orthogonal subspaces such that one set of components was similarly active during both action and imagery, and others were only active during a single task typeâ€"action or imagery. Although they inhabited orthogonal neural dimensions, the action-unique and imagery-unique subspaces contained a strikingly similar set of dynamical features. Our results suggest that during motor imagery, motor cortex maintains the same overall population dynamics as during execution by recreating the missing components related to motor output and/or feedback within a unique imagery-only subspace.

Cognitive neuroscience: Mental replay in monkeys.

Cognition allows sensory experiences to inform later actions in flexible ways. A new study shows that, in a cognitively demanding task, monkeys store visual information in short-term memory and replay it when they need it to make a decision.

How learning unfolds in the brain: toward an optimization view.

How do changes in the brain lead to learning? To answer this question, consider an artificial neural network (ANN), where learning proceeds by optimizing a given objective or cost function. This "optimization framework" may provide new insights into how the brain learns, as many idiosyncratic features of neural activity can be recapitulated by an ANN trained to perform the same task. Nevertheless, there are key features of how neural population activity changes throughout learning that cannot be readily explained in terms of optimization and are not typically features of ANNs. Here we detail three of these features: (1) the inflexibility of neural variability throughout learning, (2) the use of multiple learning processes even during simple tasks, and (3) the presence of large task-nonspecific activity changes. We propose that understanding the role of these features in the brain will be key to describing biological learning using an optimization framework.

Monkeys exhibit a paradoxical decrease in performance in high-stakes scenarios.

In high-stakes situations, people sometimes exhibit a frustrating phenomenon known as "choking under pressure." Usually, we perform better when the potential payoff is larger. However, once potential rewards get too high, performance paradoxically decreases-we "choke." Why do we choke under pressure? An animal model of choking would facilitate the investigation of its neural basis. However, it could be that choking is a uniquely human occurrence. To determine whether animals also choke, we trained three rhesus monkeys to perform a difficult reaching task in which they knew in advance the amount of reward to be given upon successful completion. Like humans, monkeys performed worse when potential rewards were exceptionally valuable. Failures that occurred at the highest level of reward were due to overly cautious reaching, in line with the psychological theory that explicit monitoring of behavior leads to choking. Our results demonstrate that choking under pressure is not unique to humans, and thus, its neural basis might be conserved across species.

Learning is shaped by abrupt changes in neural engagement.

Internal states such as arousal, attention and motivation modulate brain-wide neural activity, but how these processes interact with learning is not well understood. During learning, the brain modifies its neural activity to improve behavior. How do internal states affect this process? Using a brain-computer interface learning paradigm in monkeys, we identified large, abrupt fluctuations in neural population activity in motor cortex indicative of arousal-like internal state changes, which we term 'neural engagement.' In a brain-computer interface, the causal relationship between neural activity and behavior is known, allowing us to understand how neural engagement impacted behavioral performance for different task goals. We observed stereotyped changes in neural engagement that occurred regardless of how they impacted performance. This allowed us to predict how quickly different task goals were learned. These results suggest that changes in internal states, even those seemingly unrelated to goal-seeking behavior, can systematically influence how behavior improves with learning.

Information about contact force and surface texture is mixed in the firing rates of cutaneous afferent neurons.

Cutaneous mechanoreceptors in our hands gather information about the objects we handle. Tactile fibers encode mixed information about contact events and object properties. Neural coding in tactile afferents is typically studied by varying a single aspect of tactile stimuli, avoiding the confounds of real-world haptic interactions. We instead record responses of small populations of dorsal root ganglia (DRG) neurons to variable tactile stimuli and find that neurons primarily respond to force, though some texture information can be detected. Tactile nerve fibers convey information about many features of haptic interactions, including the force and speed of contact, as well as the texture and shape of the objects being handled. How we perceive these object features is relatively unaffected by the forces and movements we use when interacting with the object. Because signals related to contact events and object properties are mixed in the responses of tactile fibers, our ability to disentangle these different components of our tactile experience implies that they are demultiplexed as they propagate along the neuraxis. To understand how texture and contact mechanics are encoded together by tactile fibers, we studied the activity of multiple neurons recorded simultaneously in the cervical DRG of two anesthetized rhesus monkeys while textured surfaces were applied to the glabrous skin of the fingers and palm using a handheld probe. A transducer at the tip of the textured probe measured contact forces as tactile stimuli were applied at different locations on the finger-pads and palm. We examined how a sample population of DRG neurons encode force and texture and found that firing rates of individual neurons are modulated by both force and texture. In particular, slowly adapting (SA) neurons were more responsive to force than texture, and rapidly adapting (RA) neurons were more responsive to texture than force. Although force could be decoded accurately throughout the entire contact interval, texture signals were most salient during onset and offset phases of the contact interval.NEW & NOTEWORTHY Cutaneous mechanoreceptors in our hands gather information about the objects we handle. Tactile fibers encode mixed information about contact events and object properties. Neural coding in tactile afferents is typically studied by varying a single aspect of tactile stimuli, avoiding the confounds of real-world haptic interactions. We instead record responses of small populations of DRG neurons to variable tactile stimuli and find that neurons primarily respond to force, though some texture information can be detected.

Stabilization of a brain-computer interface via the alignment of low-dimensional spaces of neural activity.

The instability of neural recordings can render clinical brain-computer interfaces (BCIs) uncontrollable. Here, we show that the alignment of low-dimensional neural manifolds (low-dimensional spaces that describe specific correlation patterns between neurons) can be used to stabilize neural activity, thereby maintaining BCI performance in the presence of recording instabilities. We evaluated the stabilizer with non-human primates during online cursor control via intracortical BCIs in the presence of severe and abrupt recording instabilities. The stabilized BCIs recovered proficient control under different instability conditions and across multiple days. The stabilizer does not require knowledge of user intent and can outperform supervised recalibration. It stabilized BCIs even when neural activity contained little information about the direction of cursor movement. The stabilizer may be applicable to other neural interfaces and may improve the clinical viability of BCIs.

Brain-computer interfaces for basic neuroscience.

Brain-computer interfaces (BCIs) provide a powerful new tool for basic neuroscience investigators. This is because the BCI approach forges a tight link between the observation of neural activity and well-controlled manipulations of neural activity, driven by the BCI user's own volition. As in all branches of science, progress in neuroscience rests on observation and manipulation. In neuroscience, our observations are typically measurements of neural activity and behavior, and our manipulations are lesions and the addition of neural activity through direct neural stimulation, which cause changes in behavior and in the activity of other neurons. A BCI links observation and manipulation directly because the participant in the experiment observes a mapping of his or her own neural activity, and through volitional control, manipulates that activity. Researchers employing the BCI approach in a basic neuroscience context have made new progress toward understanding the neural basis of motor control, learning, and cognition. To date, most of the basic research using the BCI approach has been applied to understanding the motor system, but future basic science research objectives using the BCI approach include the neural basis of cognitive and emotional function, and explorations of the computational limits of neural circuitry.

A Deep Dive to Illuminate V4 Neurons.

Neurons in the primate cortical area V4 have been recognized as being selective for intermediate-level visual features, including shape, texture, and color. But, there is no a priori reason why V4 neurons should respond to easily describable categories of visual stimuli. In a recent study Bashivan et al. (Science, 2019) used an artificial neural network to synthesize images that acted as 'super-stimuli' for recorded neurons in V4, pushing the boundaries of known neural responses in this brain region and raising broader questions about the potential for synergies between biological and artificial neural networks.

New neural activity patterns emerge with long-term learning.

Learning has been associated with changes in the brain at every level of organization. However, it remains difficult to establish a causal link between specific changes in the brain and new behavioral abilities. We establish that new neural activity patterns emerge with learning. We demonstrate that these new neural activity patterns cause the new behavior. Thus, the formation of new patterns of neural population activity can underlie the learning of new skills.