Detection and evaluation of bursts in terms of novelty and surprise

Junji Ito; Emanuele Lucrezia; Günther Palm; Sonja Grün; Junji Ito; Emanuele Lucrezia; Günther Palm; Sonja Grün

doi:10.3934/mbe.2019351

Mathematical Biosciences and Engineering

2019, Volume 16, Issue 6: 6990-7008. doi: 10.3934/mbe.2019351

Previous Article Next Article

Research article Special Issues

Detection and evaluation of bursts in terms of novelty and surprise

1.
Institute for Neuroscience and Medicine (INM-6), Jülich Research Centre, 52425 Jülich, Germany
2.
Institute for Advanced Simulation (IAS-6) and Jara Brain Institute for Brain Structure and Function (INM-10), Jülich Research Centre, 52425 Jülich, Germany
3.
Institute of Neural Information Processing, University of Ulm, James-Franck-Ring, 89081 Ulm, Germany
4.
Theoretical Systems Neurobiology, RWTH Aachen University, Worringerweg 3, 52056 Aachen, Germany

Received: 28 February 2019 Accepted: 18 July 2019 Published: 31 July 2019

The detection of bursts and also of response onsets is often of relevance in understanding neurophysiological data, but the detection of these events is not a trivial task. We build on a method that was originally designed for burst detection using the so-called burst surprise as a measure. We extend this method and provide a proper significance measure. Our method consists of two stages. In the first stage we model the neuron's interspike interval (ISI) distribution and make an i.i.d. assumption to formulate our null hypothesis. In addition we define a set of 'surprising' events that signify deviations from the null hypothesis in the direction of 'burstiness'. Here the so-called (strict) burst novelty is used to measure the size of this deviation. In the second stage we determine the significance of this deviation. The (strict) burst surprise is used to measure the significance, since it is the negative logarithm of the significance probability. After showing the consequences of a non-proper null hypothesis on burst detection performance, we apply the method to experimental data. For this application the data are divided into a period for parameter estimation to express a proper null hypothesis (model of the ISI distribution), and the rest of the data is analyzed by using that null hypothesis. We find that assuming a Poisson process for experimental spike data from motor cortex is rarely a proper null hypothesis, because these data tend to fire more regularly and thus a gamma process is more appropriate. We show that our burst detection method can be used for rate change onset detection, because a deviation from the null hypothesis detected by (strict) burst novelty also covers an increase of firing rate.
- burst detection,
- interspike intervals,
- significance,
- gamma process,
- response onset detection
Citation: Junji Ito, Emanuele Lucrezia, Günther Palm, Sonja Grün. Detection and evaluation of bursts in terms of novelty and surprise[J]. Mathematical Biosciences and Engineering, 2019, 16(6): 6990-7008. doi: 10.3934/mbe.2019351

Related Papers:

Abstract

The detection of bursts and also of response onsets is often of relevance in understanding neurophysiological data, but the detection of these events is not a trivial task. We build on a method that was originally designed for burst detection using the so-called burst surprise as a measure. We extend this method and provide a proper significance measure. Our method consists of two stages. In the first stage we model the neuron's interspike interval (ISI) distribution and make an i.i.d. assumption to formulate our null hypothesis. In addition we define a set of 'surprising' events that signify deviations from the null hypothesis in the direction of 'burstiness'. Here the so-called (strict) burst novelty is used to measure the size of this deviation. In the second stage we determine the significance of this deviation. The (strict) burst surprise is used to measure the significance, since it is the negative logarithm of the significance probability. After showing the consequences of a non-proper null hypothesis on burst detection performance, we apply the method to experimental data. For this application the data are divided into a period for parameter estimation to express a proper null hypothesis (model of the ISI distribution), and the rest of the data is analyzed by using that null hypothesis. We find that assuming a Poisson process for experimental spike data from motor cortex is rarely a proper null hypothesis, because these data tend to fire more regularly and thus a gamma process is more appropriate. We show that our burst detection method can be used for rate change onset detection, because a deviation from the null hypothesis detected by (strict) burst novelty also covers an increase of firing rate.

References

[1]	K. D. Harris, H. Hirase, X. Leinekugel, et al., Temporal interaction between single spikes and complex spike bursts in hippocampal pyramidal cells, Neuron, 32 (2001), 141–149.
[2]	O. Avila-Akerberg and M. J. Chacron, Nonrenewal spike train statistics: causes and functional consequences on neural coding, Exp. Brain Res., 210 (2011), 353–371.
[3]	F. Zeldenrust, W. J. Wadman and B. Englitz, Neural coding with bursts–current state and future perspectives, Front. Comput. Neurosci., 12 (2018), 48.
[4]	A. Kepecs and J. Lisman, Information encoding and computation with spikes and bursts, Network-Comp. Neural, 14 (2003), 103–118.
[5]	D. A. Butts, P. O. Kanold and C. J. Shatz, A burst-based "hebbian" learning rule at retinogeniculate synapses links retinal waves to activity-dependent refinement, PLoS Biol., 5 (2007), e61.
[6]	J. Cocatre-Zilgien and F. Delcomyn, Identification of bursts in spike trains, J. Neurosci. Method., 41 (1992), 19–30.
[7]	D. C. Tam, An alternate burst analysis for detecting intra-burst firings based on inter-burst periods, Neurocomputing, 44 (2002), 1155–1159.
[8]	M. Chiappalone, A. Novellino, I. Vajda, et al., Burst detection algorithms for the analysis of spatio-temporal patterns in cortical networks of neurons, Neurocomputing, 65 (2005), 653–662.
[9]	B. Gourévitch and J. J. Eggermont, A nonparametric approach for detection of bursts in spike trains, J. Neurosci. Method., 160 (2007), 349–358.
[10]	L. Chen, Y. Deng, W. Luo, et al., Detection of bursts in neuronal spike trains by the mean inter-spike interval method, Prog. Nat. Sci., 19 (2009), 229–235.
[11]	V. Pasquale, S. Martinoia and M. Chiappalone, A self-adapting approach for the detection of bursts and network bursts in neuronal cultures, J. Comput. Neurosci., 29 (2010), 213–229.
[12]	D. Ko, C. Wilson, C. Lobb, et al., Detection of bursts and pauses in spike trains, J. Neurosci. Method., 211 (2012), 145–158.
[13]	I. A. Välkki, K. Lenk, J. E. Mikkonen, et al., Network-wide adaptive burst detection depicts neuronal activity with improved accuracy, Front. Comput. Neurosci., 11 (2017), 40.
[14]	D. J. Bakkum, M. Radivojevic, U. Frey, et al., Parameters for burst detection, Front. Comput. Neurosci., 7 (2014), 193.
[15]	C. R. Legendy and M. Salcman, Bursts and recurrences of bursts in the spike trains of sponta-neously active striate cortex neurons, J. Neurophysiol., 53 (1985), 926–939.
[16]	G. Palm, Evidence, information and surprise, Biol. Cybern., 42 (1981), 57–68.
[17]	S. Grün and S. Rotter (eds.), Analysis of Parallel Spike Trains, Springer, 2010.
[18]	S. Grün, M. Diesmann and A. Aertsen, 'Unitary Events' in multiple single-neuron spiking activity. I. Detection and significance, Neural Comput., 14 (2002), 43–80.
[19]	S. Grün, M. Diesmann and A. Aertsen, 'Unitary Events' in multiple single-neuron spiking activity. II. Non-Stationary data, Neural Comput., 14 (2002), 81–119.
[20]	E. Torre, D. Picado-Mui˜ no, M. Denker, et al., Statistical evaluation of synchronous spike patterns extracted by frequent item set mining, Front. Comput. Neurosci., 7 (2013), 132,
[21]	P. Quaglio, A. Yegenoglu, E. Torre, et al., Detection and evaluation of spatio-temporal spike patterns in massively parallel spike train data with spade, Front. Comput. Neurosci., 11 (2017), 41.
[22]	G. Palm, Novelty, Information and Surprise, Springer Science & Business Media, 2012.
[23]	M. R. DeWeese, M. Wehr and A. M. Zador, Binary spiking in auditory cortex, J. Neurosci., 23 (2003), 7940–7949.
[24]	H. Câteau and A. Reyes, Relation between single neuron and population spiking statistics and effects on network activity, Phys. Rev. Lett., 96 (2006), 058101.
[25]	T. Brochier, L. Zehl, Y. Hao, et al., Massively parallel recordings in macaque motor cortex during an instructed delayed reach-to-grasp task, Scientific Data, 5 (2018), 180055,
[26]	S. Tokdar, P. Xi, R. C. Kelly, et al., Detection of bursts in extracellular spike trains using hidden semi-markov point process models, J. Comput. Neurosci., 29 (2010), 203–212.
[27]	A. Riehle, S. Wirtssohn, S. Grün, et al., Mapping the spatio-temporal structure of motor cortical LFP and spiking activities during reach-to-grasp movements, Front. Neural Circuit., 7 (2013), 48,
[28]	E. Torre, P. Quaglio, M. Denker, et al., Synchronous spike patterns in macaque motor cortex during an instructed-delay reach-to-grasp task, J. Neurosci., 36 (2016), 8329–8340,
[29]	A. Riehle, T. Brochier, M. Nawrot, et al., Behavioral context determines network state and vari-ability dynamics in monkey motor cortex, Front. Neural Circuit., 12 (2018), 52.
[30]	M. P. Nawrot, C. Boucsein, V. Rodriguez Molina, et al., Measurement of variability dynamics in cortical spike trains, J. Neurosci. Method., 169 (2008), 374–390.
[31]	M. P. Nawrot, Analysis and interpretation of interval and count variability in neural spike trains, in Analysis of Parallel Spike Trains (eds. S. Rotter and S. Grün), Springer, Berlin, 2010.
[32]	C. R. Legendy, Three principles of brain function and structure, Int. J. Neurosci., 6 (1975), 237–254.
[33]	S. Shinomoto, H. Kim, T. Shimokawa, et al., Relating neuronal firing patterns to functional differ-entiation of cerebral cortex, PLOS Comput. Biol., 5 (2009), e1000433.
[34]	Y. Mochizuki, T. Onaga, H. Shimazaki, et al., Similarity in neuronal firing regimes across mam-malian species, J. Neurosci., 36 (2016), 5736–5747.
[35]	G. Maimon and J. A. Assad, Beyond poisson: Increased spike-time regularity across primate parietal cortex, Neuron, 62 (2009), 426–440.
[36]	F. Farkhooi, M. F. Strube-Bloss and M. P. Nawrot, Serial correlation in neural spike trains: Ex-perimental evidence, stochastic modeling, and single neuron variability, Phys. Rev. E, 79 (2009), 021905.
[37]	J. Csicsvari, H. Hirase, A. Czurko, et al., Reliability and state dependence of pyramidal cell–interneuron synapses in the hippocampus: an ensemble approach in the behaving rat, Neuron, 21 (1998), 179–189.
[38]	S. N. Baker and G. L. Gerstein, Determination of response latency and its application to normal-ization of cross-correlation measures, Neural Comput., 13 (2001), 1351–1377.
[39]	M. Nawrot, A. Aertsen and S. Rotter, Elimination of response latency variability in neuronal spike trains, Biol. Cybern., 5 (2003), 321–334.
[40]	A. Bollimunta, K. H. Knuth and M. Ding, Trial-by-trial estimation of amplitude and latency vari-ability in neuronal spike trains, J. Neurosci. Method., 160 (2007), 163–170.
[41]	M. Messer, M. Kirchner, J. Schiemann, et al., A multiple filter test for the detection of rate changes in renewal processes with varying variance, Ann. Appl. Stat., 8 (2014), 2027–2067.
[42]	M. Levakova, M. Tamborrino, S. Ditlevsen, et al., A review of the methods for neuronal response latency estimation, Biosystems, 136 (2015), 23–34.

Reader Comments

Your name:*

Email:*
© 2019 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)