Model-based decoupling of evoked and spontaneous neural activity in calcium imaging data

Triplett, Marcus A.; Pujic, Zac; Sun, Biao; Avitan, Lilach; Goodhill, Geoffrey J.

doi:10.1101/691261

Cited by 3 publications

(14 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the model we have proposed, we assumed that for any two stimulus values there exist sufficiently large subpopulations such that their components contribute independent and identically distributed noise when exposed to either of the two stimulus values. Now, this is certainly a simplification and a more realistic assumption would be that in the absence of sensory stimulation the activity in this subpopulations is governed by low-dimensional dynamics in addition to some individual noise [ 32 , 33 ]. In principle though, our conclusions should hold true irrespectively following the same general argument that we formalised in this work: In a large neural population, stochastic variability, even if it is only in a relatively small subpopulation, is sufficient to produce unique samples of neural activity in an experiment.…”

Section: Discussionmentioning

confidence: 99%

Limitations to Estimating Mutual Information in Large Neural Populations

Mölter¹,

Goodhill²

2020

Entropy

Self Cite

View full text Add to dashboard Cite

Information theory provides a powerful framework to analyse the representation of sensory stimuli in neural population activity. However, estimating the quantities involved such as entropy and mutual information from finite samples is notoriously hard and any direct estimate is known to be heavily biased. This is especially true when considering large neural populations. We study a simple model of sensory processing and show through a combinatorial argument that, with high probability, for large neural populations any finite number of samples of neural activity in response to a set of stimuli is mutually distinct. As a consequence, the mutual information when estimated directly from empirical histograms will be equal to the stimulus entropy. Importantly, this is the case irrespective of the precise relation between stimulus and neural activity and corresponds to a maximal bias. This argument is general and applies to any application of information theory, where the state space is large and one relies on empirical histograms. Overall, this work highlights the need for alternative approaches for an information theoretic analysis when dealing with large neural populations.

show abstract

Section: Discussionmentioning

confidence: 99%

Limitations to Estimating Mutual Information in Large Neural Populations

Mölter¹,

Goodhill²

2020

Entropy

Self Cite

View full text Add to dashboard Cite

show abstract

“…However, while solving such optimisation problems is central to obtaining usable scientific data from raw calcium imaging recordings, few models have addressed the subsequent stage of analysis, which focuses on relating the segmented, denoised calcium signals to experimentally relevant variables using principled statistical methods. Several recent examples include statistical modelling of calcium responses evoked by sensory stimuli [169], internal factors [170], and photostimulation [128]. In ref.…”

Section: Introductionmentioning

confidence: 99%

“…In ref. [170], Triplett et al focus on separating the calcium influx evoked by sensory stimuli from spontaneous calcium transients attributable to hidden internal factors. This is done by assuming an additive interaction between evoked and spontaneous activity, and optimising how calcium influx is assigned to each source under a sparse prior on internal factor activity (see below for details).…”

Section: Introductionmentioning

confidence: 99%

“…First, recent work [169] has noted that, because a substantial fraction of calcium influx is exactly zero, spike-and-slab distributions such as the zero-inflated gamma are better suited to modelling calcium influx compared to the exponential prior used in ref. [170]. Second, in ref.…”

Section: Introductionmentioning

confidence: 99%

“…Second, in ref. [170] the evoked response is modelled deterministically with respect to the presented stimulus, so that any variability in stimulus responses arises either through addition with ongoing spontaneous activity or through a Gaussian imaging noise model. Third, due to the linearity of the model components required to perform the decoupling in ref.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Neural encoding models for multivariate optical imaging data

Triplett¹

View full text Add to dashboard Cite

The nervous system must integrate information arriving via peripheral sensory pathways with internal factors that regulate circuit function, cognitive state, and behaviour. This thesis introduces new methodology for modelling such latent internal factors in calcium imaging data, for characterising how they give rise to structured patterns of neural activity, and for understanding their role in neural development. We construct two statistical models that bridge factor analysis methods for neural spike train data with statistical signal processing algorithms for calcium imaging data. We apply our techniques to calcium imaging of zebrafish optic tectum and mouse visual cortex, demonstrating their ability to decompose large scale optical recordings of neural activity into low dimensional factors that encode sensory stimulation, spontaneous activity, and modulation of multiplicative excitability. We then use computational modelling to explore how such latent structure could arise via synaptic plasticity in denervated neural circuits, identifying a Hebbian learning rule that allows neural circuits to self-organise into a highly modular state where assemblies of neurons activate spontaneously. Together, this thesis work provides practical open source tools for probabilistic modelling of optical imaging data, and insight into the mechanisms underlying the development of structured neural activity. i I would like to thank my teachers and mentors at the University of Auckland, especially Patrick Girard, André Nies, Cristian Calude, and Jeremy Seligman. You all were an inspiration to me during my formative undergraduate years. Finally, my most heartfelt thanks go to my family for their unending enthusiasm in my seemingly unending academic pursuits. vi

show abstract

Information processing in the developing zebrafish brain

Moelter¹

View full text Add to dashboard Cite

Every action in response to some environmental state is preceded by a cascade of activity in the brain. Information about the environment is perceived and enters the brain through sensory systems, such as the visual, auditory, or somatosensory system, is processed as it travels downstream along the particular sensory pathway and at the end may elicit for example a motor response. In order to achieve this, the brain is made up of a multitude of neurons, ranging from less than one hundred in simple organisms to several billion cells in great apes. These discrete units of neural signalling facilitate the propagation of information through electrochemical processes. As the animal develops and gains experience about the environment these circuits and the neural code, the representation of information in neural activity, undergo changes that likely allow the animal to act more efficiently. Key insights into the functioning of the brain can be gained from studying its development. And in that sense understanding the development of the neural code can give new insight into how the brain processes information. In order to do so, we turned to the larval zebrafish as a model system and in this system studied the development of neural activity in the optic tectum. The optic tectum is a brain structure that is homologous to the superior colliculus in the mammalian brain and an integral part of the visual systems in vertebrates such as fish. In recent years, the zebrafish has become a popular model system in systems neuroscience because of the interesting feature that in their larval stage zebrafish are translucent. This together with their small size allows for in vivo microscopy and using genetically encoded calcium indicators this enables non-invasive recording of neural activity from large neural populations at single cell resolution. Contributions by others to this thesisGeoffrey J. Goodhill and Lilach Avitan have provided support and guidance in the analysis and interpretation of data as well as feedback on the writing. The calcium imaging experiments including the preprocessing of the calcium imaging recordings were performed by Lilach Avitan, Zac Pujic, Biao Sun, Shuyu Zhu, Matthew Van De Poll, and Ann-Elin Myhre. The images of zebrafish shown were taken by Zac Pujic. Statement of parts of this thesis submitted to qualify for the award of another degreeNo works submitted towards another degree have been included in this thesis. Research involving human or animal subjectsAll animal procedures were approved by The University of Queensland Ethics Committee, Approval Number QBI/152/16/ARC. The approval certificate is included in Appendix A.

show abstract

Model-based decoupling of evoked and spontaneous neural activity in calcium imaging data

Cited by 3 publications

References 35 publications

Limitations to Estimating Mutual Information in Large Neural Populations

Limitations to Estimating Mutual Information in Large Neural Populations

Neural encoding models for multivariate optical imaging data

Information processing in the developing zebrafish brain

Contact Info

Product

Resources

About