Database Download

Stereo-EEG recordings database

2014-04-30T16:10:17Z

Stereo-EEG recordings database of epileptic seizures

The database include three set of data: intracranial Stereo-EEG (SEEG) recordings, adjacency matrices and indexes derived from graph theory

1) SEEG recordings consisted of 3 minutes of interictal activity from subjects with intractable focal seizures. The data were recorded during the invasive pre-surgical work-up at the Epilepsy Surgery center of Niguarda Hospital in Milan. Subjects were monitored for up to several days in order to characterize their seizures and assess their candidacy for surgery. Recordings were collected from 10 subjects with type II focal cortical dysplasia. Raw signals were sampled at 1000 Hz and, after low pass filtering at 120 Hz to avoid aliasing, downsampled at 250 Hz. The number of channels included in each record ranged from 70 to 90. In order to protect the privacy of the subjects, all health information in the original files was removed.

2) Adjacency matrix: EEG recordings was divided into 36 non-overlapped epochs (5 s length) and each couple of channels was submitted to connectivity analysis by means of nonlinear regression method h2. An adjacency matrix was then built for each epoch, with the value of the element ij representing the association or ﬂow of information from channel xj(t) to xi(t).

3) From each adjacency matrix we extracted 11 indexes aimed at quantified the centrality of the nodes: degree, strength, betweenness centrality; network segregation: clustering coefficient, local efficiency, modularity, and network integration: path length, efficiency.

For each patients, three different file mat were provided:
- PtxxSigBip.mat : File containing the EEG signals in bipolar montage
- PtxxAdMatr.mat : File containing 36 Adjacency Matrices
- PtxxPar.mat: File containing 36 matrices with the 11 graph parameters estimated for each EEG channel.

The following .zip files include all the data just described.

SEEG part 1 SEEG part 2 SEEG part 3 SEEG part 4 SEEG part 5
SEEG part 6 SEEG part 7 SEEG part 8 SEEG part 9 SEEG part 10

The following .zip includes the detailed description of the database.

Database Description

Please reference the following publications:

- D.Y. Takahashi, L. A. Baccal, K. Sameshima, Connectivity inference between neural structrures via partial directed coherence, Journal of Applied Statistics, 34(10), pp. 1259-1273.

- F. Panzica, G. Varotto, F. Rotondi, R. Spreafico, S., Identification of the epileptogenic zone from stereo-EEG signals: a connectivity-graph theory approach, Frontiers in Neurology, 4.

- M. Rubinov, O. Sporns, Complex netwrok measurs of the brain connectivity: uses and interpretations, Nueroimage, 52(3), pp. 1059-1069.

- G. Varotto, L. Tasso, S. Franceschetti, R. Spreafico, F. Panzica, Epileptogenic networks of type II focal cortical dysplasia: a stereo-EEG study, NeuroImage, 61(3), pp.591-598.

- F. Wending, F. Bartolomei, J. Bellanger, P.Chauvel, Interpretation of interdependencies in epileptic signals using a macroscopic physiological model of the EEG, Clinical neurophysiology, 112(7), pp. 1201-1218.

fMRI - EEG database of epileptic seizure

2014-04-30T14:03:10Z

fMRI - EEG database of epileptic seizure

The database contains two types of recordings collected as part the ACTIVE project at Tel-Aviv Sourasky Medical Center (TASMC) Israel.

The first set is an FMRI dataset including language FMRI testing and anatomical MRI that were part of a machine learning language lateralization project. These patients have also undergone the invasive WADA test for language lateralization. The second set is EEG-fMRI recording for the detection of the epileptic network. This dataset includes both EEG and fMRI recordings, recorded simultaneously and anatomical MRI. This dataset also includes the markings of at least one trained epileptologist specifying the timings in which the epileptologist found epileptic activity. Both datasets were collected from epilepsy patients candidates for neurosurgical resection of their epileptogenic foci.

In the following file you can find the description of the database:

Epilepsy Dataset Description fMRI-EEG

To retrieve the content of the described material, please contact: tomergazit@gmail.com

Semiology of epileptic seizure database

2014-04-28T06:17:13Z

Semiology of epileptic seizure database

The clinical dataset of patients that we used in our work was obtained from three clinical epileptological centers: Carlo Besta Neurological Institute, Niguarda Ca Granda Hospital and San Paolo Hospital, all in Milan, Italy. All patients are drug-resistant epileptic patient, who do not present any other pathology (e.g. malignant tumors), and they were operated on and seizure-free for at least one year. All the patients have signed an informed consent to the treatment of their anonymized data.

The dataset is generated as follows. Once clinicians have a collection of video recordings of epileptic seizures, they create a table for each patient in which 29 symptoms are present and which gives the duration in seconds of the seizures. The first 14 rows of the table correspond to the subjective symptoms and the next 15 to the objective symptoms. For each patient, clinicians fill in the table, assigning to the symptoms observed a '1', and a '0' to manifestation not observed. The Semiology Database includes 79 unlabeled patients, of which: 50 patients present temporal lobe epilepsy and 29 patients with extra-temporal lobe epilepsy. The second file (i.e. Key Table Localization) indicates the correct classification of patients and the hospital origin.

For a detailed description of the semiology of epileptic seizure database, please have a look at http://www.sciencedirect.com/science/article/pii/S0933365714000207

DATABASE