Data Analysis

The primary aim of this line of research is the development of measures of spike train synchrony. So far, most emphasis has been put on the ISI- and the SPIKE-distance (confer references below). All our measures are first tested and compared to previously published measures in a controlled setting in which they are applied to simulated data with well-known properties. In a second step they are applied to real neuronal data kindly provided by collaborating laboratories.

So far we have analyzed data from four different laboratories located on the extended University of California at San Diego (UCSD) campus in La Jolla, CA, USA: the Institute for Nonlinear Science (INLS, UCSD), the Gentner songbird lab (UCSD), the Reynolds visual attention lab (Salk), and the Chichilnisky retina lab (Salk). All this data complement each other in that they fulfill different requirements and are suited to address different issues of neuronal dynamics. Further data have been provided by Florian Mormann from the Department of epilspsy at the University of Bonn, Germany.

At INLS Julie Haas (now at Harvard) recorded in vitro cortical cells from the medial entorhinal cortex of young Long-Evans rats. These data have been used to investigate the influence of synaptic inputs on steadily spiking neurons and to develop a minimum Markov model based on a new spike classification scheme (2010 paper). Furthermore, we used them to illustrate both the ISI-distance (2007 paper) and its extensions (2009 paper).

Further data have been provided by Jude Mitchell from the Reynolds lab at the Salk Institute. Jude records in vivo single unit data of area V4 in a Macaque monkey during a visual attention task. These data offer higher statistics (many repeated stimulations producing a large number of spikes) and were thus perfectly suited to address the issue of multi-neuron spike train variability. They were already employed in the 2009 paper to illustrate the advantages of our instantaneous measures.

Another line of research is performed in collaboration with the Gentner songbird lab in the Department of Psychology at UCSD. In this lab Emily Caporello, a PhD student in the UCSD neuroscience program, records neuronal single- and multi-unit responses to song stimuli from the auditory system of songbirds. These data offer the advantage of having been evoked by stimuli that change on different time scales (i.e., the notes, syllables and motifs of a birdsong) and they are thus optimal to address the temporal aspects of spike train analysis. Currently they are used to investigate potential limitations of time-scale dependent spike train distances.

More recently we have focused on the analysis of retina data kindly provided by Martin Greschner from the Chichilnisky lab at the Salk institute. Chichilnisky's lab uses a state-of-the-art 512-electrode recording system that allows to monitor hundreds of cells at once while stimulating the retina with spatio-temporal light patterns. In our 2011 paper we used one of these datasets to compare the performance of several single-neuron spike train distances. Due to the large numbers of neurons recorded in parallel these data are also tailor-made to test population spike train distances.

At the department of epilepsy at the University of Bonn Florian Mormann records neuronal spiking from the human medial temporal lobe of epilepsy patients undergoing seizure monitoring prior to epilepsy surgery. This is one of the most prominent setups in which a real-time application of measures of spike train syanchrony would be very desirable. Thus in our 2013 paper these data were used to illustrate the vastly improved SPIKE-distance and its real-time variant.


“In trecordings of neuronal spiking from thehuman medial temporal lobe. These recordings were performedat the University of Bonn in epilepsy patients undergoingseizure monitoring prior to epilepsy surgery.his example the ISI-Distance is applied to two simulated Hindemarsh-Rose time series (for more details see the methods section)



Kreuz T, Haas JS, Morelli A, Abarbanel HDI, Politi A: Measuring spike train synchrony. J Neurosci Methods 165, 151 (2007) [PDF]

Kreuz T, Chicharro D, Andrzejak RG, Haas JS, Abarbanel HDI: Measuring multiple spike train synchrony. J Neurosci Methods 183, 287 (2009) [PDF]

Haas JS*,Kreuz T*, Torcini A, Politi A, Abarbanel HDI: Rate maintenance and resonance in the entorhinal cortex. Eur J Neurosci 32, 1930 (2010) [PDF]

Kreuz T, Chicharro D, Greschner M, Andrzejak RG: Time-resolved and time-scale adaptive measures of spike train synchrony. J Neurosci Methods 195, 92 (2011) [PDF]

Chicharro D, Kreuz T, Andrzejak RG: What can spike train distances tell us about the neural code? J Neurosci Methods 199, 146 (2011) [PDF]

Andrzejak RG, Kreuz T: Characterizing unidirectional couplings between point processes and flows. European Physics Letters 96, 50012 (2011) [PDF]

Kreuz T: Measures of neuronal signal synchrony. Scholarpedia 6(12), 11922 (2011) [HTML]

Kreuz T: Measures of spike train synchrony. Scholarpedia 6(10), 11934 (2011) [HTML]

Houghton C, Kreuz T: On the efficient calculation of van Rossum distances. Network: Computation in neural systems 23, 48 (2012) [PDF]

Kreuz T: SPIKE-distance. Scholarpedia 7(12), 30652 (2012) [HTML]

Kreuz T, Chicharro D, Houghton C, Andrzejak RG, Mormann F: Monitoring spike train synchrony. J Neurophysiol. In press (2012). Also submitted to the arXiv [PDF]