Independent coding of absolute duration and distance magnitudes in the prefrontal cortex

Same principle, but different computations in representing time and space

Time and space are two intertwined contexts that frame our cognition of the world and have shared mechanisms. A well-known theory on this case is "A Theory of Magnitude (ATOM)" which states that the perception of these two domains shares common mechanisms. However, evidence regarding shared computations of time and space is intermixed. To investigate this issue, we asked human subjects to reproduce time and distance intervals with saccadic eye movements in similarly designed tasks. We applied an observer model to both modalities and found underlying differences in the processing of time and space. While time and space computations are both probabilistic, adding priors to space perception minimally improved model performance, as opposed to time perception which was consistently better explained by Bayesian computations. We also showed that while both measurement and motor variability were smaller in distance than time reproduction, only the motor variability was correlated between them, as both tasks used saccadic eye movements for response. Our results suggest that time and space perception abide by the same algorithm but have different computational properties.

Categorical representation of abstract spatial magnitudes in the executive telencephalon of crows

The ability to group abstract continuous magnitudes into meaningful categories is cognitively demanding but key to intelligent behavior. To explore its neuronal mechanisms, we trained carrion crows to categorize lines of variable lengths into arbitrary "short" and "long" categories. Single-neuron activity in the nidopallium caudolaterale (NCL) of behaving crows reflected the learned length categories of visual stimuli. The length categories could be reliably decoded from neuronal population activity to predict the crows' conceptual decisions. NCL activity changed with learning when a crow was retrained with the same stimuli assigned to more categories with new boundaries ("short", "medium," and "long"). Categorical neuronal representations emerged dynamically so that sensory length information at the beginning of the trial was transformed into behaviorally relevant categorical representations shortly before the crows' decision making. Our data show malleable categorization capabilities for abstract spatial magnitudes mediated by the flexible networks of the crow NCL.

Social monitoring of actions in the macaque frontopolar cortex

The frontopolar cortex (FPC) of primates appeared as a main innovation in the evolution of anthropoid primates and it has been placed at the top of the prefrontal hierarchy. The only study to date that investigated the activity of FPC neurons in monkeys performing a cognitive task suggested that these cells were involved in the monitoring of self-generated actions. We recorded the activity of neurons in the FPCs of two rhesus monkeys while they performed a social variant of a nonmatch-to-goal task that required monitoring the actions of a human or computer agent. We discovered that the role of FPC neurons extends beyond self-generated actions to include monitoring others' actions. Their monitoring activity was very specific. First, neurons in the FPC encoded the spatial position of the target but not its object features. Second, a dedicated representation of the human agent actions was tied to the time of target acquisition, while it was reduced or absent in the successive epochs of the trial. Finally, this other-specific neural substrate did not emerge during the interaction with a virtual agent such as the computer. These results provide a new perspective on the functions of a uniquely primate brain area, suggesting that FPC might play an important role in social behaviors.

Partially Separable Aspects of Spatial and Temporal Estimations in Virtual Navigation as Revealed by Adaptation

Recent studies claim that estimating the magnitude of the spatial and temporal aspects of one's self-motion shows similar characteristics, suggesting shared processing mechanisms between these two dimensions. While the estimation of other magnitude dimensions, such as size, number, and duration, exhibits negative aftereffects after prolonged exposure to the stimulus, it remains to be elucidated whether this could occur similarly in the estimation of the distance travelled and time elapsed during one's self-motion. We sought to fill this gap by examining the effects of adaptation on distance and time estimation using a virtual navigation task. We found that a negative aftereffect occurred in the distance reproduction task after repeated exposure to self-motion with a fixed travel distance. No such aftereffect occurred in the time reproduction task after repeated exposure to self-motion with a fixed elapsed time. Further, the aftereffect in distance reproduction occurred only when the distance of the adapting stimulus was fixed, suggesting that it did not reflect adaptation to time, which varied with distance. The estimation of spatial and temporal aspects of self-motion is thus processed by partially separable mechanisms, with the distance estimation being similar to the estimation of other magnitude dimensions.

The role of neural tuning in quantity perception

Perception of quantities, such as numerosity, timing, and size, is essential for behavior and cognition. Accumulating evidence demonstrates neurons processing quantities are tuned, that is, have a preferred quantity amount, not only for numerosity, but also other quantity dimensions and sensory modalities. We argue that quantity-tuned neurons are fundamental to understanding quantity perception. We illustrate how the properties of quantity-tuned neurons can underlie a range of perceptual phenomena. Furthermore, quantity-tuned neurons are organized in distinct but overlapping topographic maps. We suggest that this overlap in tuning provides the neural basis for perceptual interactions between different quantities, without the need for a common neural representational code.

Neuronal activity in the monkey prefrontal cortex during a duration discrimination task with visual and auditory cues

To investigate neuronal processing involved in the integration of auditory and visual signals for time perception, we examined neuronal activity in prefrontal cortex (PFC) of macaque monkeys during a duration discrimination task with auditory and visual cues. In the task, two cues were consecutively presented for different durations between 0.2 and 1.8 s. Each cue was either auditory or visual and was followed by a delay period. After the second delay, subjects indicated whether the first or the second cue was longer. Cue- and delay-responsive neurons were found in PFC. Cue-responsive neurons mostly responded to either the auditory or the visual cue, and to either the first or the second cue. The neurons responsive to the first delay showed activity that changed depending on the first cue duration and were mostly sensitive to cue modality. The neurons responsive to the second delay exhibited activity that represented which cue, the first or second cue, was presented longer. Nearly half of this activity representing order-based duration was sensitive to cue modality. These results suggest that temporal information with visual and auditory signals was separately processed in PFC in the early stage of duration discrimination and integrated for the final decision.

Slower prefrontal metastable dynamics during deliberation predicts error trials in a distance discrimination task

Cortical activity related to erroneous behavior in discrimination or decision-making tasks is rarely analyzed, yet it can help clarify which computations are essential during a specific task. Here, we use a hidden Markov model (HMM) to perform a trial-by-trial analysis of the ensemble activity of dorsolateral prefrontal cortex (PFdl) neurons of rhesus monkeys performing a distance discrimination task. By segmenting the neural activity into sequences of metastable states, HMM allows us to uncover modulations of the neural dynamics related to internal computations. We find that metastable dynamics slow down during error trials, while state transitions at a pivotal point during the trial take longer in difficult correct trials. Both these phenomena occur during the decision interval, with errors occurring in both easy and difficult trials. Our results provide further support for the emerging role of metastable cortical dynamics in mediating complex cognitive functions and behavior.

Hidden Markov Models Predict the Future Choice Better Than a PSTH-Based Method

Beyond average firing rate, other measurable signals of neuronal activity are fundamental to an understanding of behavior. Recently, hidden Markov models (HMMs) have been applied to neural recordings and have described how neuronal ensembles process information by going through sequences of different states. Such collective dynamics are impossible to capture by just looking at the average firing rate. To estimate how well HMMs can decode information contained in single trials, we compared HMMs with a recently developed classification method based on the peristimulus time histogram (PSTH). The accuracy of the two methods was tested by using the activity of prefrontal neurons recorded while two monkeys were engaged in a strategy task. In this task, the monkeys had to select one of three spatial targets based on an instruction cue and on their previous choice. We show that by using the single trial's neural activity in a period preceding action execution, both models were able to classify the monkeys' choice with an accuracy higher than by chance. Moreover, the HMM was significantly more accurate than the PSTH-based method, even in cases in which the HMM performance was low, although always above chance. Furthermore, the accuracy of both methods was related to the number of neurons exhibiting spatial selectivity within an experimental session. Overall, our study shows that neural activity is better described when not only the mean activity of individual neurons is considered and that therefore, the study of other signals rather than only the average firing rate is fundamental to an understanding of the dynamics of neuronal ensembles.

Retrospective time estimation following damage to the prefrontal cortex

Time estimation in patients with prefrontal cortex (PFC) damage is often inaccurate. The relationship between PFC and estimation of short time intervals has been examined. However, it remains unclear whether PFC damage affects estimation of longer time intervals. Here, we investigated the ability of patients and healthy subjects to verbally estimate a period of 30 min, using a method easily applied in clinical settings. In 99 patients with brain damage, we compared under and normal ranges of time in patients with PFC damage or damage to other brain areas with the chi-squared test. Subsequently, we conducted a discriminant analysis and a multiple linear regression analysis to identify specific brain areas affecting time estimation. We observed a significantly larger number of patients who overestimated 30 min in the group with bilateral PFC damage compared to patients with damage to other regions. Discriminant analysis revealed that damage of right lateral PFC and left medial PFC contributed to discrimination between the normal range and overestimation groups. Multiple linear regression analysis indicated that right lateral PFC damage strongly affected overestimation of a 30-min interval. Neuropsychological test results revealed lower general cognitive function scores and orientation scores in overestimation group. The length of estimated time and the score of delayed word recall were negatively correlated. We propose that these may require encoding, maintenance, and updating of memory and are indirectly related to contextual memory. We discuss hypotheses on contextual memory segmentation and reconstruction to clarify the mechanism of impaired time overestimation in PFC-damaged patients.

Outcome Modulation Across Tasks in the Primate Dorsolateral Prefrontal Cortex

Animals need to learn and to adapt to new and changing environments so that appropriate actions that lead to desirable outcomes are acquired within each context. The prefrontal cortex (PF) is known to underlie such function that directly implies that the outcome of each response must be represented in the brain for behavioral policies update. However, whether such PF signal is context dependent or it is a general representation beyond the specificity of a context is still unclear. Here, we analyzed the activity of neurons in the dorsolateral PF (PFdl) recorded while two monkeys performed two perceptual magnitude discrimination tasks. Both tasks were well known by the monkeys and unexpected changes did not occur but the difficulty of the task varied from trial to trial and thus the monkeys made mistakes in a proportion of trials. We show a context-independent coding of the response outcome with neurons maintaining similar selectivity in both task contexts. Using a classification method of the neural activity, we also show that the trial outcome could be well predicted from the activity of the same neurons in the two contexts. Altogether, our results provide evidence of high degree of outcome generality in PFdl.

Interference between Space and Time Estimations: From Behavior to Neurons

Influences between time and space can be found in our daily life in which we are surrounded by numerous spatial metaphors to refer to time. For instance, when we move files from one folder to another in our computer a horizontal line that grows from left to right informs us about the elapsed and remaining time to finish the procedure and, similarly, in our communication we use several spatial terms to refer to time. Although with some differences in the degree of interference, not only space has an influence on time but both magnitudes influence each other. Indeed, since our childhood our estimations of time are influenced by space even when space should be irrelevant and the same occurs when estimating space with time as distractor. Such interference between magnitudes has also been observed in monkeys even if they do not use language or computers, suggesting that the two magnitudes are tightly coupled beyond communication and technology. Imaging and lesion studies have indicated that same brain areas are involved during the processing of both magnitudes and have suggested that rather than coding the specific magnitude itself the brain represents them as abstract concepts. Recent neurophysiological studies in prefrontal cortex, however, have shown that the coding of absolute and relative space and time in this area is realized by independent groups of neurons. Interestingly, instead, a high overlap was observed in this same area in the coding of goal choices across tasks. These results suggest that rather than during perception or estimation of space and time the interference between the two magnitudes might occur, at least in the prefrontal cortex, in a subsequent phase in which the goal has to be chosen or the response provided.

A Bayesian Perspective on Accumulation in the Magnitude System

Several theoretical and empirical work posit the existence of a common magnitude system in the brain. Such a proposal implies that manipulating stimuli in one magnitude dimension (e.g. duration in time) should interfere with the subjective estimation of another magnitude dimension (e.g. size in space). Here, we asked whether a generalized Bayesian magnitude estimation system would sample sensory evidence using a common, amodal prior. Two psychophysical experiments separately tested participants on their perception of duration, surface, and numerosity when the non-target magnitude dimensions and the rate of sensory evidence accumulation were manipulated. First, we found that duration estimation was resilient to changes in surface and numerosity, whereas lengthening (shortening) the duration yielded under- (over-) estimations of surface and numerosity. Second, the perception of surface and numerosity were affected by changes in the rate of sensory evidence accumulation, whereas duration was not. Our results suggest that a generalized magnitude system based on Bayesian computations would minimally necessitate multiple priors.