Dimensionality reduction for large-scale neural recordings

Cunningham, John P.; Yu, Byron M.

doi:10.1038/nn.3776

Cited by 1,018 publications

(1,001 citation statements)

References 91 publications

(167 reference statements)

Supporting

Mentioning

985

Contrasting

Order By: Relevance

“…Therefore, it is plausible that insights into the function of sensory areas can be gained from application of modelling techniques with reduced assumptions on the computational properties of neuronal networks. Recent research on latent variable models provides the mathematical techniques for such models with limited assumptions [7]. One biologically plausible model, rectified latent variable model (RLVM), has been shown to accurately predict the activity of neurons in the rat barrel cortex [11].…”

mentioning

confidence: 99%

Weighted Network Density Predicts Range of Latent Variable Model Accuracy

Palmerston

She

Chan

2018

Preprint

View full text Add to dashboard Cite

Abstract-Current experimental techniques impose spatial limits on the number of neuronal units that can be recorded in-vivo. To model the neural dynamics utilizing these sampled data, Latent Variable Models (LVMs) have been proposed to study the common unobserved processes within the system that drives neural activities, through an implicit network with hidden states. Yet, relationships between these latent variable models and widely-studied network connectivity measures remained unclear. In this paper, a biologically plausible latent variable model was first fit to neural activity recorded via 2-photon microscopic calcium imaging in the murine primary visual cortex. Graph theoretic measures were then applied to quantify network properties in the recorded sub-regions. Comparison of weighted network measures with LVM prediction accuracy shows some network measures having a strong relationship with LVM prediction accuracy, while other measures do not have a robust relationship with LVM prediction accuracy. Results show LVM will achieve high accuracy in dense networks. I. INTRODUCTIONSince the late 1950s, analyses of in-vivo single-unit recordings in the primary visual cortex have focused on correlating activity of individual neurons with observable features of visual stimuli [8]. However, not all neurons can be confidently classified as maximally responding to a single stimulus feature. Anatomically, it is known that primary sensory regions receive inputs from areas not involved with the stimuli of interest [4]. Imaging techniques have made it possible to record the activity of many neurons simultaneously, in-vivo, while an animal is awake [6]. Applying graph theoretic tools on these high-volume simultaneous recordings allows for novel analyses of brain networks. Graph theory can efficiently characterize network types and elucidate important network properties [1]. For example, in the study of sensory perception, this confluence of techniques allows for a change from pairwise associations between individual sensory neurons and sensory stimuli, to network-level measures of sensory representation. Conventional graph-theoretic measures have often evaluated the explicit network where nodes and edges correspond to neuronal units recorded and statistical dependences between these neurons respectively [1]. Yet, the explicit neuronal population dynamics could be modulated by latent variables or implicit network with unobservable nodes and edges, where the implicit activities were not recorded [5]. Therefore, it is plausible that insights into the function of sensory areas can be gained from application of modelling techniques with reduced assumptions on the computational properties of neuronal networks. Recent research on latent variable models provides the mathematical techniques for such models with limited assumptions [7]. One biologically plausible model, rectified latent variable model (RLVM), has been shown to accurately predict the activity of neurons in the rat barrel cortex [11]. Through simulation and applicatio...

show abstract

mentioning

confidence: 99%

Weighted Network Density Predicts Range of Latent Variable Model Accuracy

Palmerston

She

Chan

2018

Preprint

View full text Add to dashboard Cite

show abstract

“…[119]). The complexity of cellular-scale whole-brain recordings presents many challenges, which can be addressed for example by reducing the dimensionality of the datasets through various clustering or component mapping techniques [120][121][122].…”

Section: Predicting Brain Dynamicsmentioning

confidence: 99%

Cerebral cartography and connectomics

Sporns

2015

Phil. Trans. R. Soc. B

View full text Add to dashboard Cite

One contribution of 11 to a theme issue 'Cerebral cartography: a vision of its future'. Cerebral cartography and connectomics pursue similar goals in attempting to create maps that can inform our understanding of the structural and functional organization of the cortex. Connectome maps explicitly aim at representing the brain as a complex network, a collection of nodes and their interconnecting edges. This article reflects on some of the challenges that currently arise in the intersection of cerebral cartography and connectomics. Principal challenges concern the temporal dynamics of functional brain connectivity, the definition of areal parcellations and their hierarchical organization into large-scale networks, the extension of whole-brain connectivity to cellular-scale networks, and the mapping of structure/function relations in empirical recordings and computational models. Successfully addressing these challenges will require extensions of methods and tools from network science to the mapping and analysis of human brain connectivity data. The emerging view that the brain is more than a collection of areas, but is fundamentally operating as a complex networked system, will continue to drive the creation of ever more detailed and multi-modal network maps as tools for on-going exploration and discovery in human connectomics.

show abstract

“…A chain of action potentials, generated by a single neuron is called a spike train and is a time sequence of firing events, which occur at regular or irregular intervals. Thus, these pulse-coded signals that represent the information encoded by a neuron, employ both binary and temporal coding mechanisms, which expand the signals into higher dimensional spaces [2].…”

Section: Introductionmentioning

confidence: 99%

Dictionary learning for spontaneous neural activity modeling

Troullinou

Tsagkatakis

Palagina

et al. 2017

2017 25th European Signal Processing Conference (EUSIPCO)

View full text Add to dashboard Cite

Abstract-Modeling the activity of an ensemble of neurons can provide critical insights into the workings of the brain. In this work we examine if learning based signal modeling can contribute to a high quality modeling of neuronal signal data. To that end, we employ the sparse coding and dictionary learning schemes for capturing the behavior of neuronal responses into a small number of representative prototypical signals. Performance is measured by the reconstruction quality of clean and noisy test signals, which serves as an indicator of the generalization and discrimination capabilities of the learned dictionaries. To validate the merits of the proposed approach, a novel dataset of the actual recordings from 183 neurons from the primary visual cortex of a mouse in early postnatal development was developed and investigated. The results demonstrate that high quality modeling of testing data can be achieved from a small number of training examples and that the learned dictionaries exhibit significant specificity when introducing noise.

show abstract

Dimensionality reduction for large-scale neural recordings

Cited by 1,018 publications

References 91 publications

Weighted Network Density Predicts Range of Latent Variable Model Accuracy

Weighted Network Density Predicts Range of Latent Variable Model Accuracy

Cerebral cartography and connectomics

Dictionary learning for spontaneous neural activity modeling

Contact Info

Product

Resources

About