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

By C. Pandarinath, D. O’Shea, Jasmine Collins, Rafal Józefowicz, S. Stavisky, J. Kao, Eric M. Trautmann, Matthew T. Kaufman, Stephen I. Ryu, Leigh R. Hochberg and 4 more. 2017

800 citations recorded by the source corpus.

Neuroscience is experiencing a revolution in which simultaneous recording of thousands of neurons is revealing population dynamics that are not apparent from single-neuron responses. This structure is typically extracted from data averaged across many trials, but deeper understanding requires studying phenomena detected in single trials, which is challenging due to incomplete sampling of the neural population, trial-to-trial variability, and fluctuations in action potential timing. We introduce latent factor analysis via dynamical systems, a deep learning method to infer latent dynamics from single-trial neural spiking data. When applied to a variety of macaque and human motor cortical datasets, latent factor analysis via dynamical systems accurately predicts observed behavioral variables, extracts precise firing rate estimates of neural dynamics on single trials, infers perturbations to those dynamics that correlate with behavioral choices, and combines data from non-overlapping recording sessions spanning months to improve inference of underlying dynamics. LFADS, a deep learning method for analyzing neural population activity, can extract neural dynamics from single-trial recordings, stitch separate datasets into a single model, and infer perturbations, for example, from behavioral choices to these dynamics.

Subfields: Cellular and Molecular Neuroscience, Cognitive Neuroscience

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cellular and Molecular Neuroscience

In Cognitive Neuroscience

Citations in the corpus

Unsupervised Discovery of Demixed, Low-Dimensional Neural Dynamics across Multiple Timescales through Tensor Component Analysis (2017, methodology): “Simple linear models have a relatively constrained dynamical repertoire [12], while models with nonlinear 520 elements often have greater predictive abilities [17, 18], but at the expense of interpretability.”
Manifold GPLVMs for discovering non-Euclidean latent structure in neural data (2020, methodology): “Many of these methods are explicitly probabilistic, with notable examples including GPFA (Yu et al., 2009) and LFADS (Pandarinath et al., 2018).”
Interpreting neural computations by examining intrinsic and embedding dimensionality of neural activity (2021, methodology): “…in this direction by developing sophisticated nonlinear dimensionality reduction techniques and applying them to large-scale neural data [10,11,14,19,49], and by using artificial neural network models to generate testable hypotheses for how neural systems might emulate complex cognitive…”
Artificial Neural Networks for Neuroscientists: A Primer (2020, background): “The norm of this matrix product, k Wur T k , can grow exponentially with T if W r is large (more precisely, the largest eigenvalue of W r > 1) or vanish to zero ifW r is small, making it historically difficult to train recurrent networks (Bengio et al., 1994; Pascanu et al., 2013).”
Parametric UMAP Embeddings for Representation and Semisupervised Learning (2020, methodology): “…networks are necessary for real-time analyses and manipulations; thus, deep neural networks are increasingly being used (Brown & De Bivort, 2018; Pandarinath et al., 2018; Sainburg, Thielk, & Gentner, 2019). n The speed of embedding is especially important in signal processing paradigms where…”
Long-term stability of cortical population dynamics underlying consistent behavior (2020, positive): “Recent work has further shown that population-based approaches can improve decoding stability 6,26,46 , enable unsupervised decoding 44 , or even adjust for changes in neural activity based on its statistics alone 45 .”
Towards the neural population doctrine. (2019, methodology): “Gao et al. [30] introduced ANNs as the nonlinear map from latent dynamics to observable neural population activity, a theme which has been elaborated both by including ANNs in the dynamics model, and otherwise [63–66].”
Tree-Structured Recurrent Switching Linear Dynamical Systems for Multi-Scale Modeling (2018, background): “In neuroscience, sequential autoencoders offer highly accurate models of brain activity (Pandarinath et al., 2018).”
On the methods for reactivation and replay analysis (2020, background): “Finally, the recent development of deep learning techniques that can capture the statistics of neuronal patterns, even in single trials [76] is an interesting direction that will probably attract a lot of attention in the coming years.”