A multi-region recurrent circuit for evidence accumulation in rats

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: July 11, 2024

Abstract Decision-making based on noisy evidence requires accumulating and categorizing it to form a choice. Here we evaluate proposed feedforward modular mapping of this process in rats: accumulated anterodorsal striatum (ADS) is categorized prefrontal cortex (frontal orienting fields, FOF). Contrary this, show that both regions appear be indistinguishable their encoding/decoding accumulator value communicate information bidirectionally. Consistent with role for FOF accumulation, silencing ADS projections impacted behavior throughout the accumulation period, even while nonselective did not. We synthesize these findings into multi-region recurrent neural network trained novel approach. In-silico experiments reveal multiple scales recurrence cortico-striatal circuit rescue computation upon perturbations. These results suggest accumulate distributed manner, yielding redundant representations robustness certain

Language: Английский

---

The neural and computational architecture of feedback dynamics in mouse cortex during stimulus report

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2023, Volume and Issue: unknown

Published: July 19, 2023

Abstract Conscious reportability of visual input is associated with a bimodal neural response in primary cortex (V1): an early-latency coupled to stimulus features and late-latency report or detection. This late wave activity, central major theories consciousness, thought be driven by prefrontal (PFC), responsible for “igniting” it. Here we analyzed two electrophysiological studies mice performing different detection tasks, characterize activity profiles three key cortical regions: V1, posterior parietal (PPC) PFC. We then developed minimal network model, constrained known connectivity between these regions, reproducing the spatio-temporal propagation visual-and report-related activity. Remarkably, while PFC was indeed necessary generate this occurred only through mediation PPC. PPC, not PFC, had final veto enabling V1

Language: Английский

---

Circuit mechanisms of navigation strategy learning in mice

Current Biology, Journal Year: 2023, Volume and Issue: 34(1), P. 79 - 91.e4

Published: Dec. 14, 2023

Navigation tasks involve the gradual selection and deployment of increasingly effective searching procedures to reach targets. The brain mechanisms underlying such complex behavior are poorly understood, but their elucidation might provide insights into systems linking exploration decision making in learning. Here, we developed a trial-by-trial goal-related search strategy analysis as mice learned navigate identical water mazes encompassing distinct rules monitored process throughout We found that navigation learning involved following three phases: an early phase during which maze-specific strategies deployed minority trials, second preferential increasing one strategy, final commitment this only. maze phases were affected differently by inhibition retrosplenial cortex (RSC), dorsomedial striatum (DMS), or dorsolateral (DLS). Through region-specific inactivation experiments gain-of-function involving activation learning-related cFos+ ensembles, unraveled how relates these sequential processes. RSC is critically important for selection, DMS mediates deployment, DLS ensures consistency Notably, specific ensembles was sufficient direct (RSC) (DMS) different maze. Our results establish approach dissect unsupervised processes suggest involves evidence-based direction RSC, reinforcement-modulated through DMS, online guidance DLS.

Language: Английский

---

Cortex-wide characterization of decision-making neural dynamics during spatial navigation

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Oct. 23, 2024

Summary Decision-making during freely moving behaviors involves complex interactions among many cortical and subcortical regions. However, the spatiotemporal coordination across regions to generate a decision is less understood. Using head-mounted widefield microscope, cortex-wide calcium dynamics were recorded in mice expressing GCaMP7f as they navigated an 8-maze using two paradigms. The first was alternating pattern that required short term memory of previous trial make correct second after rule change fixed path which rewards delivered only on left side. Identification activation states revealed differences between There higher probability for visual/retrosplenial state paradigm secondary motor posterior parietal left-only. Three sequences (motifs) illustrated both anterior activity propagations cortex. propagating motifs had highest around peaked following decision. latter, likely reflecting internal feedback influence future actions, more common left-only paradigm. Therefore, probabilities differ when working versus trajectory reward

Language: Английский

---

Determination of Neuronal Activity and Its Meaning for the Processes of Learning and Memory

Advances in neurobiology, Journal Year: 2024, Volume and Issue: unknown, P. 3 - 38

Published: Jan. 1, 2024

Language: Английский

---

Strain Gauge Measuring System for Subsensory Micromotions Analysis as an Element of a Hybrid Human–Machine Interface

Sensors, Journal Year: 2022, Volume and Issue: 22(23), P. 9146 - 9146

Published: Nov. 25, 2022

The human central nervous system is the integrative basis for functioning of organism. such integration provided by fact that same neurons are involved in various sets sensory, cognitive, and motor functions. Therefore, analysis one set components makes it possible to draw conclusions about state efficiency other components. Thus, evaluate a person’s cognitive properties, we can assess their involuntary acts, i.e., subsensory reactions. To measure parameters have developed strain gauge measuring system. This provides measurement estimation movements against background voluntary isometric efforts. article presents architecture shows organization primary signal processing analog form, particular separation taken from strain-gauge sensor into frequency smoothly varying averaging subtracting signals. transfer form simplifies implementation digital part allowed minimizing response time while displaying forces visual feedback channel. describes realization elements results its experimental research.

Language: Английский

---

Methylphenidate modulates motor cortical dynamics and behavior

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2023, Volume and Issue: unknown

Published: Oct. 17, 2023

Abstract Methylphenidate (MPH, brand: Ritalin) is a common stimulant used both medically and non-medically. Though typically prescribed for its cognitive effects, MPH also affects movement. While it known that noncompetitively blocks the reuptake of catecholamines through inhibition dopamine norepinephrine transporters, critical step in exploring how behavior to understand directly neural activity. This would establish an electrophysiological mechanism action MPH. Since we now have biologically-grounded network-level hypotheses regarding populations motor cortical neurons plan execute movements, there unique opportunity make testable predictions systemic administration – pharmacological perturbation might affect activity cortex. To end, administered clinically-relevant doses Rhesus monkeys as they performed instructed-delay reaching task. Concomitantly, measured from dorsal premotor primary Consistent with our predictions, found dose-dependent significant effects on reaction time, trial-by-trial variability, movement speed. We confirmed changes time variability were accompanied by previously established population-level preparatory condition-independent signal precedes movements. expected speed be result amplitude dynamics and/or translation those space. Instead, data are consistent whereby neuromodulatory effect increase gain signal-to-noise during reaching. Continued work this domain better brain-wide other psychoactive drugs could facilitate more targeted treatments host cognitive-motor disorders.

Language: Английский

---

Curriculum learning inspired by behavioral shaping trains neural networks to adopt animal-like decision making strategies

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Jan. 15, 2024

Recurrent neural networks (RNN) are ubiquitously used in neuroscience to capture both dynamics and behaviors of living systems. However, when it comes complex cognitive tasks, traditional methods for training RNNs can fall short capturing crucial aspects animal behavior. To address this challenge, we take inspiration from a commonly (though rarely appreciated) approach the experimental neuroscientist's toolkit: behavioral shaping. Our solution leverages task compositionality models animal's relevant learning experiences prior task. Taking as target temporal wagering previously studied rats, designed pretraining curriculum simpler tasks that prerequisites performing well. These not just simplified versions task, but reflect sub-computations. We show is required adopt similar strategies including long-timescale inference latent states, which conventional approaches fail capture. Mechanistically, our supports development key dynamical systems features needed implementing value-based decision making. Overall, addresses gap network model by incorporating inductive biases animals, important modeling rely on computational abilities acquired past experiences.

Language: Английский

---

Prefrontal cortex encodes behavior states decoupled from movement

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 12, 2024

Abstract Prefrontal cortex is often viewed as an extension of the motor system, but little understood how it relates to natural behavior. We therefore tracked kinematics freely moving rats performing minimally structured tasks and measured which aspects behavior were read out in prefrontal neural populations. Naturalistic behaviors such rearing or chasing a bait each encoded by unique ensembles, behavioral representations not anchored posture movement. Rather, coding kinematic features depended on their relevance animals’ current task animal performed. Behavior-specific ensembles preceded outlasted physical actions and, accordingly, population activity evolved at slower timescales than cortex. These findings argue that locked output, may instead reflect motivations perform certain rather themselves. Highlights uniquely encode different naturalistic Behavioral tuning explained movement Population evolves M1 Single-cell varies across yet ensemble stable

Language: Английский

---

An optical brain-machine interface reveals a causal role of posterior parietal cortex in goal-directed navigation

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 30, 2024

Abstract Cortical circuits contain diverse sensory, motor, and cognitive signals, form densely recurrent networks. This creates challenges for identifying causal relationships between neural populations behavior. We developed a calcium imaging-based brain-machine interface (BMI) to study the role of posterior parietal cortex (PPC) in controlling navigation virtual reality. By training decoder estimate navigational heading velocity from PPC activity during navigation, we discovered that mice could immediately navigate toward goal locations when control was switched BMI. No learning or adaptation observed BMI, indicating naturally occurring patterns are sufficient drive trajectories real time. During successful BMI trials, decoded decoupled mouse’s physical movements, suggesting relates intended trajectories. Our work demonstrates offers approach investigating links

Language: Английский

---