Temple University EEG Corpus

To request access to the TUH EEG Corpus, please fill out this form and email a signed copy to help@nedcdata.org. Please include "Download The TUH EEG Corpus" in the subject line or click on this link.

Note this is an Adobe Acrobat form, and it is best filled out using Adobe Acrobat or a similarly compatible tool. We suggest you download a copy to your desktop and fill it out using a local app, rather than attempt to complete the form from within a browser.

The form must be filled out correctly or it will be returned to you. Please follow the instructions on the form very carefully, including completing the address information accurately. This is very important and we cannot accept forms with incorrect addresses.

Once your form is accepted, you will receive information about the credentials used to access the data in a separate email, and be added to our listserv. This usually takes about 24 to 48 hours. We need to track who downloads the data. We also want to be able to inform you of any updates to the releases.

Corpora

Once you have successfully registered and transmitted your ssh keys, you can download our corpora using rsync. The path for the most current release is shown with each entry.

The TUH EEG Corpus (TUEG) [rsync path: data/tuh_eeg/tuh_eeg/v2.0.1]: A rich archive of 26,846 clinical EEG recordings collected at Temple University Hospital (TUH) from 2002 - 2017. Read this journal paper for a more complete description of the corpus.
The TUH Abnormal EEG Corpus (TUAB) [rsync path: data/tuh_eeg/tuh_eeg_abnormal/v3.0.1]: A corpus of EEGs that have been annotated as normal or abnormal. Read Silvia Lopez's MS thesis for a description of the corpus.
The TUH EEG Artifact Corpus (TUAR) [rsync path: data/tuh_eeg/tuh_eeg_artifact/v3.0.1]: This subset of TUEG that contains annotations of 5 different artifacts: (1) eye movement (EYEM), (2) chewing (CHEW), (3) shivering (SHIV), (4) electrode pop, electrode static, and lead artifacts (ELPP), and (5) muscle artifacts (MUSC).
The TUH EEG Epilepsy Corpus (TUEP) [rsync path: data/tuh_eeg/tuh_eeg_epilepsy/v2.0.1]: This is a subset of TUEG that contains 100 subjects epilepsy and 100 subjects without epilepsy, as determined by a certified neurologist. The data was developed in collaboration with a number of partners including NIH.
The TUH EEG Events Corpus (TUEV) [rsync path: data/tuh_eeg/tuh_eeg_events/v2.0.1]: This corpus is a subset of TUEG that contains annotations of EEG segments as one of six classes: (1) spike and sharp wave (SPSW), (2) generalized periodic epileptiform discharges (GPED), (3) periodic lateralized epileptiform discharges (PLED), (4) eye movement (EYEM), (5) artifact (ARTF) and (6) background (BCKG).
The TUH EEG Seizure Corpus (TUSZ) [rsync path: data/tuh_eeg/tuh_eeg_seizure/v2.0.3]: This corpus has EEG signals that have been manually annotated data for seizure events (start time, stop, channel and seizure type). For more information about this corpus, please refer to our book section. The electrode configurations and annotation guidelines are described here.
The TUH EEG Slowing Corpus (TUSL) [rsync path: data/tuh_eeg/tuh_eeg_slowing/v2.0.1]: This is another subset of TUEG that contains annotations of slowing events. This corpus has been used to study common error modalities in automated seizure detection.

Software

This section of our web site contains a rich collection of machine learning systems and supporting software, including our best EEG seizure detection system.

NEDC EEG Annotation System (EAS: v5.1.2): A tool that allows rapid annotation of EEG signals. The tool includes spectrogram and energy plots, and is capable of transcribing data in real time. Learn more about this tool from our IEEE SPMB 2018 paper.
NEDC ResNet Decoder Real-Time (ERDR: v1.0.1): A real-time EEG seizure detection system based on a ResNet-18 neural network and transfer learning. This package contains a real-time decoder that is described in this publication. This is also part of our real-time demonstration system.
NEDC ResNet Seizure Detection System (EEGR: v1.0.1): A research version of our real-time seizure detection system that includes a trainer and a decoder. This package is used to develop and evaluate models. This package contains a decoder that is described in this publication.
NEDC Eval EEG (EVAL v5.0.0): A Python-based scoring package that implements a variety of standard evaluation metrics. A complete description of the software can be found here.
NEDC PyPrint EDF (PYPR v1.0.0): A Python-based tool that decodes the header and signal data in an Edf file. This is not as simple as it might seem because channels can be permuted and must be decoded using channel labels.
MATLAB EDF (MEDF): MATLAB code that loads EEG signal data from an EDF file.
EDF Browser (EDFB): An open-source program that can be used to view files such as EEG, EMG, ECG, etc., available for Windows and Linux.
Python-based EDF (PYED): A Python interface to EDFLib that lets you read and write EDF files (the distribution format for TUH EEG).

Documentation

We have several tutorials available to facilitate acquisition of our data, including these videos.

Electrodes (ELEC): A document that describes how EEG signals are stored in a multichannel signal file format. This document also includes a description of the channel labels, which are required to properly decode the data.
Annotations: (ANNO): A document that describes how we annotate seizures and store the annotations in various file formats.
Reports: (RPRT): We no longer distribute reports with our corpora. However, we do offer users the opportunity to search the anonymized reports for keywords or items of interest. If you provide a simple Python script that executes your search criteria, we can run it on the reports, and report which files matched your search criteria.

A example script demonstrating the interface we need is here. This script takes as input a file containing a list of filenames, and returns filenames that match your search criteria.

Our users have used this service to select subsets of our corpora that meet their needs. Note, however, that searching these reports is not trivial, since the reports are unstructured text. It is very difficult to parse these using natural language processing tools to avoid false positives when doing these kinds of searchers.

Users interested in this service should email a simple Python script that does not use any non-standard libraries to help@nedcdata.org.

Instructions

As of January 2026, our released corpora are now distributed using ssh keys and rsync. The process for this begins with your submission of the above form. Once approved, you will receive instructions on how to transmit your key and access the data.

Rsync, which is available on Linux and Mac platforms, is our preferred way of downloading data. It allows you to easily keep your copy of the data in sync with ours. Windows users can get access to rsync by installing MobaXterm. Some tips on how to install and use MobaXterm are here. Before you attempt to download an entire corpus, you should test your ability to download data by executing this command:

rsync -auvxL -e "ssh -i ~/.ssh/id_ed25519" \
nedc-tuh-eeg@www.isip.piconepress.com:data/tuh_eeg/TEST .

This command must be typed on one line in your command line tool.

If for some reason this fails, change "-auxvL" to "-auxvvvL". This will generate a log file that your IT support team can use to diagnose the problems with your downloads.

Note that the "-L" option in rsync instructs it to follow links. All of our corpora are linked back to TUEG. It is best to always use the "-L" option.

If Internet connectivity is a problem, you can send us a 8T USB drive. We will copy the data to this disk and send it to you. You must arrange for postage as described below. If you elect this option, you need to send us a 8T USB drive and provide a UPS or FedEx account number for return shipping.

Please send us a conventional USB-mounted disk drive. We have had problems with other types of media such as thumb drives. Any standard USB-powered USB 2.0 or 3.0 compatible 8T drive will work fine. Because of the time it takes to copy the data, we need a drive that can maintain a stable connection, and thumb drives have proven to be unreliable.

Mail the drive to:

Joseph Picone
1610 Rhawn Street
Philadelphia, PA 19111
Tel: 708-848-2846

Please email us for details before shipping the drive. If you ship us a drive directly from a reseller such as Amazon, please make sure that the shipment contains information that we can use to identify you. This information should include a point of contact (POC), the name of your institution, and contact information (name, surface mail address and telephone number for the POC).

Please note that disk drives sent to international destinations will often get caught in Customs for weeks. Rsync is a much better option than going through your local governments.

If you are having trouble deciding what to do, email us and describe what specific resources in which you are interested. We will be happy to guide you through the process.