Release Announcement: IBM TUSZ Pre-Processed Data (v1.0.0)

[Sean Ferrell]

Hello everyone,

We have released v1.0.0 of IBM TUSZ Pre-Processed Data (IBMPPD). This dataset was prepared by IBM Research - Australia and is available at:

https://www.isip.piconepress.com/projects/tuh_eeg/downloads/ibm_seizure_preprocessed/

v1.0.0 of IBMPPD represents a preprocessed version of the TUH Seizure Detection Corpus. The preprocessing is performed using two methods, both of which use an FFT sliding window approach (STFT) in the beginning.

In method one, the FFT magnitudes are logarithmically (base 10) calculated for the frequency bins. These bin values are then collected from all channels and a 1-D vector is created. In method two, the FFT values are normalized across frequency buckets and correlation coefficients are calculated. The eigenvalues are calculated from this correlation matrix. The eigenvalues and correlation matrix's upper triangle are used to generate features.

To learn more about this dataset, please read the following paper:

https://arxiv.org/abs/1902.01012

As always, don't hesitate to contact us with your questions or bug reports.

Best regards,

Sean Ferrell

NEDC Data Resources Development Manager
The Neural Engineering Data Consortium
Temple University
College of Engineering, ENGR 602
1947 North 12th Street
Philadelphia, PA 19122-6018
Tel: 215-204-4841
Fax: 215-204-5960

Email: sean.f...@temple.edu