Software & Data

Progress in Neuroscience is to a large extent limited by the fact that results are hard to be replicated. For example we can come up with million different spike sorting algorithms but this is not useful unless we can compare them with the same data sets.

Our policy is to share our codes (hoping they will be useful to other researchers as they are to ourselves) and some of our data (for example, all the simulations we used to test our clustering).

All the following codes and data are given for free. The only thing we ask in return is that you cite the original reference if you use them.


Unsupervised spike detection and sorting of extracellular multiunit recordings. It uses wavelets and Super-paramagnetic clustering. The code is described in:
Unsupervised spike detection and sorting with wavelets and superparamagnetic clustering.



Realistic simulation of extracellular recordings using detailed neuron models

neurocube web










Supplementary Material

A previous software for realistic simulations of extracellular recordings (Martinez et al., J Neurosci Methods, 2009)


Automatic Denoising of single-trial evoked potentials from the background EEG. It is based on wavelet multi-resolution decomposition. The code is described in:

Automatic Denoising of Single-Trial Evoked Potentials

Codes, Tutorial, and Sample Data




A method for extracting information from time patterns using wavelets and information theory. Given a raster of spike trains, WI estimates the information in the spike timing (compared to total spike counts), denoise the spike trains and estimate correlations between neurons. You can dowload the codes and exemplar data (, the help file (tutorial.PDF) and the original reference below.

Visit the WI dedicated page for downloads.


WI screen shot


Calculates the cross-correlation, and non-linear interdependence of bivariate signals.

This program was use in the paper:
Performance of different synchronization measures in real data: a case study on EEG signals.

and with coupled dynamical systems in:
Learning driver-response relationships from synchronization patterns.



Given a 1 column input ascii file, calculates its Windowed Fourier Transform and then calculates the power spectra and the Shannon
and Kulback-Leibler (relative) entropies. It also calculates the  relative band intensity ratio (RIR).

Download file: spect.m

This software was used for calculating the RIR in:
Searching for Hidden Information with Gabor Transform in Generalized Tonic-Clonic Seizures.

The windowed Fourier transform in:
Frequency evolution during tonic-clonic seizures.

The Shannon and Kulback-Leibler entropies in:
Kullback-Leibler and Renormalized Entropy: Applications to EEGs of Epilepsy Patients.


Dataset # 1: Human single-cell recording

This example is a 30' multiunit recording in the temporal lobe of an epileptic patient from Itzhak Fried's lab at UCLA. Try wave_clus for spike detection and sorting of this data.

Download file (74Mb .zip)

Dataset # 2: Simulated extracellular recordings

These simulated datasets are described in:
Unsupervised spike detection and sorting with wavelets and superparamagnetic clustering.

where they were used for testing the spikes sorting algorithm. You should be able to reproduce the results of the paper and, eventually, compare wave_clus with other spike sorting algorithms.

Dataset # 3: Simulated extracellular recordings

These simulated datasets are described in:
Realistic simulation of extracellular recordings.

this dataset contains five simulations of 2' including the activity of one multi-unit and two single-units for different firing rates and signal-to-noise ratio levels.

Simulation 1 (22Mb)
Simulation 2 (22Mb)
Simulation 3 (22Mb)
Simulation 4 (22Mb)
Simulation 5 (22Mb)

Dataset # 4: EEG signals from rats

Each example contains 5 sec of a two-channel EEG recording at left and right frontal cortex of male adult WAG/Rij rats. Signals were referenced to an electrode placed at the cerebelum, they were filtered between 1-100 Hz and digitized at 200 Hz. Example A correspond to a normal EEG and examples B. C, D and E contain spike-wave discharges.

We thank Giles van Luijtelaar and Joyce Welting for allowing their distribution.

Download files:






Details on the recordings as well as on the physiological results can be obtained from:
The reticular thalamic nucleus is involved in left-right EEG synchronization.

the first datasets were analyzed in the paper:
Performance of different synchronization measures in real data: a case study on electroencephalographic signals.

and the last two were also analyzed in:
Event synchronization: a simple a fast method to measure synchronicity and time delay patterns.

Dataset # 5: Pattern visual evoked potentials

Each file corresponds to the recording on a different subject in the left occipital electrode (O1), with linked earlobes reference. Each file contains several artifact-free trials, each of them containing 512 data points (256 pre- and 256 post-stimulation) stored with a sampling frequency of 250 Hz. Trials are stored consecutively in a 1 column file. Data was pre-filtered in the range 0.1-70Hz. All trials correspond to target stimulation with an oddball paradigm (see the papers for details).

We thanks Martin Schuermann for allowing their distribution.

Download files:

cg_olt.asc (30 trials) 
ja_olt.asc (16 trials)

These files were used for showing a denoising implementation that helps to visualize single-trial evoked potentials in:
Obtaining single  trial EPs with wavelet denoising.

Moreover, we show a time-frequency analysis of evoked potentials using Wavelets in:
Functions and sources of evoked EEG alpha oscillations studied with the Wavelet Transform

and in:

Wavelet Transform in the analysis of the frequency composition of evoked potentials.

and we show the application of an entropy defined from the distribution of wavelet coefficients in:
Wavelet-entropy: a measure of order in evoked potentials.

and in:
Wavelet entropy in event-related potentials: a new method shows frequency tuning of EEG-oscillations.

Dataset # 6: Tonic-clonic (Grand Mal) seizures

These files show tonic-clonic seizures of two subjects recorded with a scalp rigth central (C4) electrode (linked earlobes reference). It contains a total of 3 minutes with about 1 min pre-seizure, the seizure and some post-seizure activity. Sampling rate is 102.4 Hz (see the papers for more details).

Download files:


A time-frequency analysis of these type of seizures was done in:
Searching for Hidden Information with Gabor Transform in Generalized Tonic-Clonic Seizures

and in:
Frequency evolution during tonic-clonic seizures.
in press

Dataset # 7: Ongoing EEG activity

Five data sets containing quasi-stationary, artifact-free EEG signals both in normal subjects and epileptic patients were put in the web by Ralph Andrzejak from the Epilepsy center in Bonn, Germany. Each data set contains 100 single channel EEG segments of 23.6 sec duration. The data can be downloaded from:

A study of non-linear determinism of this data has been published in Phys. Rev. E (follow the link from the previous address).

Share this page:

Contact Us

Centre for Systems Neuroscience

Centre for Medicine,
Department of Neuroscience, Psychology and Behaviour
University of Leicester
15 Lancaster Rd,
Leicester LE1 7HA


T +44 (0)116 252 3249


Directions from the train station