Kernel density compression for real-time Bayesian encoding/decoding of unsorted hippocampal spikes

Sodkomkham, Danaipat; Ciliberti, Davide; Wilson, Matthew A.; Fukui, Kenichi; Moriyama, Koichi; Numao, Masayuki; Kloosterman, Fabian

doi:10.1016/j.knosys.2015.09.013

Cited by 26 publications

(33 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The readout of the content of memory reactivation during rest and slow wave sleep (SWS) is conventionally carried out in an offline analysis (Davidson et al, 2009;Pfeiffer and Foster, 2013;Roumis and Frank, 2015;Gomperts et al, 2015). An ''online'' extension of ''place''-decoding analysis has been proposed using a Bayesian spike-sorting-free encoding and decoding framework (Chen et al, 2012;Kloosterman et al, 2014;Deng et al, 2015;Sodkomkham et al, 2016).…”

Section: Introductionmentioning

confidence: 99%

Real-Time Readout of Large-Scale Unsorted Neural Ensemble Place Codes

Ciliberti

Grosmark

et al. 2018

Cell Reports

Self Cite

View full text Add to dashboard Cite

Graphical AbstractHighlights d Spike-sorting-free decoding reconstructs the rat's position with ultrafast speed d GPU-powered population decoding significantly speeds up multi-core CPU-based system d GPU computing empowers real-time assessment of decoded ''memory replay'' candidates d Open-source software toolkit supports closed-loop contenttriggered intervention SUMMARY Uncovering spatial representations from large-scale ensemble spike activity in specific brain circuits provides valuable feedback in closed-loop experiments. We develop a graphics processing unit (GPU)-powered population-decoding system for ultrafast reconstruction of spatial positions from rodents' unsorted spatiotemporal spiking patterns, during run behavior or sleep. In comparison with an optimized quad-core central processing unit (CPU) implementation, our approach achieves an $20to 50-fold increase in speed in eight tested rat hippocampal, cortical, and thalamic ensemble recordings, with real-time decoding speed (approximately fraction of a millisecond per spike) and scalability up to thousands of channels. By accommodating parallel shuffling in real time (computation time <15 ms), our approach enables assessment of the statistical significance of onlinedecoded ''memory replay'' candidates during quiet wakefulness or sleep. This open-source software toolkit supports the decoding of spatial correlates or content-triggered experimental manipulation in closed-loop neuroscience experiments.

show abstract

Section: Introductionmentioning

confidence: 99%

Real-Time Readout of Large-Scale Unsorted Neural Ensemble Place Codes

Ciliberti

Grosmark

et al. 2018

Cell Reports

Self Cite

View full text Add to dashboard Cite

show abstract

“…In recent years, marked point process models have become increasingly common in the analysis of population neural spiking activity [1][2][3]. For multi-unit spike data, these models directly relate the occurrences of spikes with particular waveform features to the biological, behavioral, or cognitive variables encoded by the population, without the need for a separate spike-sorting step.…”

Section: Introductionmentioning

confidence: 99%

“…For this reason, these are sometimes called 'clusterless' neural models. Clusterless models have been shown to capture coding properties for spikes that cannot be sorted with confidence and to lead to improved population decoding results in place field data from rat hippocampus during spatial navigation tasks [1][2][3]. Additionally, avoiding a computationally intensive spike-sorting step allows for neural decoding to be implemented in real-time, closed-loop experiments.…”

Section: Introductionmentioning

confidence: 99%

Assessing Goodness-of-Fit in Marked-Point Process Models of Neural Population Coding via Time and Rate Rescaling

Amidi

Yousefi

Nazari

et al. 2020

Preprint

View full text Add to dashboard Cite

Marked-point process models have recently been used to capture the coding properties of neural populations from multi-unit electrophysiological recordings without spike sorting. These 'clusterless' models have been shown in some instances to better describe the firing properties of neural populations than collections of receptive field models for sorted neurons and to lead to better decoding results. To assess their quality, we previously proposed a goodness-of-fit technique for markedpoint process models based on time-rescaling, which for a correct model, produces a set of uniform samples over a random region of space. However, assessing uniformity over such a region can be challenging, especially in high dimensions.Here, we propose a set of new transformations both in time and in the space of spike waveform features, which generate events that are uniformly distributed in the new mark and time spaces. These transformations are scalable to multi-dimensional mark spaces and provide uniformly distributed samples in hypercubes, which are well suited for uniformity tests. We discuss properties of these transformations and demonstrate aspects of model fit captured by each transformation. We also compare multiple uniformity tests to determine their power to identify lack-of-fit in the rescaled data. We demonstrate an application of these transformations and Ground Intensity

show abstract

“…An extension of the point process framework being increasingly applied to neural data analysis, the marked point process model [17], which can account for instantaneous joint spiking events [18], and has also allowed characterization of spikes according to waveform and the modeling of the populations to be done in single, integrated step. This approach, sometimes known as clusterless encoding and decoding [19,20,21,22], where a joint model for the firing intensity dependent on waveform and stimulus features, behavioral variables or in hippocampal recordings, spatial position, the joint mark intensity function (JMIF) is estimated simultaneously in one step, allowing a streamlined real-time analysis pipeline.…”

Section: Introductionmentioning

confidence: 99%

Marked point process filter for clusterless and adaptive encoding-decoding of multiunit activity

Arai

Liu

Frank

et al. 2018

Preprint

View full text Add to dashboard Cite

Real-time, closed-loop experiments can uncover causal relationships between specific neural activity and behavior. An important advance in realizing this is the marked point process filtering framework which utilizes the "mark" or the waveform features of unsorted spikes, to construct a relationship between these features and behavior, which we call the encoding model. This relationship is not fixed, because learning changes coding properties of individual neurons, and electrodes can physically move during the experiment, changing waveform characteristics. We introduce a sequential, Bayesian encoding model which allows incorporation of new information on the fly to adapt the model in real time. A possible application of this framework is to the decoding of the contents of hippocampal ripples in rats exploring a maze. During physical exploration, we observe the marks and positions at which they occur, to update the encoding model, which is employed to decode contents of ripples when rats stop moving, and switch back to updating the model once the rat starts moving again.

show abstract

Kernel density compression for real-time Bayesian encoding/decoding of unsorted hippocampal spikes

Cited by 26 publications

References 41 publications

Real-Time Readout of Large-Scale Unsorted Neural Ensemble Place Codes

Real-Time Readout of Large-Scale Unsorted Neural Ensemble Place Codes

Assessing Goodness-of-Fit in Marked-Point Process Models of Neural Population Coding via Time and Rate Rescaling

Marked point process filter for clusterless and adaptive encoding-decoding of multiunit activity

Contact Info

Product

Resources

About