Biology subjects

Ryu, S. I.

Publications and source records attributed to Ryu, S. I..

3 recordsLinked to original sources

bioRxiv2018

Population dynamics of choice representation in dorsal premotor and primary motor cortex

Studies in multiple species have revealed the existence of neural signals that lawfully co-vary with different aspects of the decision-making process, including choice, sensory evidence that supports the choice, and reaction time. These signals, often interpreted as the representation of a decision variable (DV), have been identified in several motor preparation circuits and provide insight about mechanisms underlying the decision-making process. However, single-trial dynamics of this process or its representation at the neural population level remain poorly understood. Here, we examine the representation of the DV in simultaneously recorded neural populations of dorsal premotor (PMd) and primary motor (M1) cortices of monkeys performing a random dots direction discrimination task with arm movements as the behavioral report. We show that single-trial DVs covary with stimulus difficulty in both areas but are stronger and appear earlier in PMd compared to M1 when the stimulus duration is fixed and predictable. When temporal uncertainty is introduced by making the stimulus duration variable, single-trial DV dynamics are accelerated across the board and the two areas become largely indistinguishable throughout the entire trial. These effects are not trivially explained by the faster emergence of motor kinematic signals in PMd and M1. All key aspects of the data were replicated by a computational model that relies on progressive recruitment of units with stable choice-related modulation of neural population activity. In contrast with several recent results in rodents, decision signals in PMd and M1 are not carried by short sequences of activity in non-overlapping groups of neurons but are instead distributed across many neurons, which once recruited, represent the decision stably during individual behavioral epochs of the trial.

neuroscience↗

bioRxiv2017

Unsupervised discovery of demixed, low-dimensional neural dynamics across multiple timescales through tensor components analysis

Perceptions, thoughts and actions unfold over millisecond timescales, while learned behaviors can require many days to mature. While recent experimental advances enable large-scale and long-term neural recordings with high temporal fidelity, it remains a formidable challenge to extract unbiased and interpretable descriptions of how rapid single-trial circuit dynamics change slowly over many trials to mediate learning. We demonstrate a simple tensor components analysis (TCA) can meet this challenge by extracting three interconnected low dimensional descriptions of neural data: neuron factors, reflecting cell assemblies; temporal factors, reflecting rapid circuit dynamics mediating perceptions, thoughts, and actions within each trial; and trial factors, describing both long-term learning and trial-to-trial changes in cognitive state. We demonstrate the broad applicability of TCA by revealing insights into diverse datasets derived from artificial neural networks, large-scale calcium imaging of rodent prefrontal cortex during maze navigation, and multielectrode recordings of macaque motor cortex during brain machine interface learning.

neuroscience↗

bioRxiv2017

Inferring single-trial neural population dynamics using sequential auto-encoders

Neuroscience is experiencing a data revolution in which simultaneous recording of many hundreds or thousands of neurons is revealing structure in population activity that is not apparent from single-neuron responses. This structure is typically extracted from trial-averaged data. Single-trial analyses are challenging due to incomplete sampling of the neural population, trial-to-trial variability, and fluctuations in action potential timing. Here we introduce Latent Factor Analysis via Dynamical Systems (LFADS), a deep learning method to infer latent dynamics from single-trial neural spiking data. LFADS uses a nonlinear dynamical system (a recurrent neural network) to infer the dynamics underlying observed population activity and to extract de-noised single-trial firing rates from neural spiking data. We apply LFADS to a variety of monkey and human motor cortical datasets, demonstrating its ability to predict observed behavioral variables with unprecedented accuracy, extract precise estimates of neural dynamics on single trials, infer perturbations to those dynamics that correlate with behavioral choices, and combine data from non-overlapping recording sessions (spanning months) to improve inference of underlying dynamics. In summary, LFADS leverages all observations of a neural populations activity to accurately model its dynamics on single trials, opening the door to a detailed understanding of the role of dynamics in performing computation and ultimately driving behavior.

neuroscience↗