Open Seizure Database v1.0.0

Citation Author(s):
Graham
Jones
Open Seizure Detector
Jamie
Pordoy
University of West London
Submitted by:
jamie pordoy
Last updated:
Thu, 08/17/2023 - 04:32
DOI:
10.21227/jftq-3e97
Data Format:
License:
0
0 ratings - Please login to submit your rating.

Abstract 

This research introduces the Open Seizure Database and Toolkit as a novel, publicly accessible resource designed to advance non-electroencephalogram seizure detection research. This paper highlights the scarcity of resources in the non-electroencephalogram domain and establishes the Open Seizure Database as the first openly accessible database containing multimodal sensor data from 49 participants in real-world, in-home environments. The database is comprised of 494 events, encompassing 146 epileptic seizures, collected over a duration of 453 days, presenting the most extensive publicly available non-electroencephalogram seizure data to date. Additionally, the database has 348 labelled false alarms, including 302 common human movements and activities. The Open Seizure Toolkit is designed to facilitate machine and deep learning practices by streamlining data from the Open Seizure Database. Utilising these resources, researchers can rapidly develop and train seizure detection models before deploying them to the Open Seizure Detector Android application. Access to these resources is expected to foster collaborative efforts, ultimately contributing to the establishment of a non-electroencephalogram gold standard and advancing the field of seizure detection.

Instructions: 

OpenSeizureDatabase

This repository contains the toolkit to create and proocess the database of seizure and seizure-like data that has been contributed by OpenSeizureDetector users to contribute to research to improve seizure detection algorithms (Open Seizure Toolkit, OSTK).

The repository contains a number of python tools that form a toolkit to assist users of the database in working with the JSON files containing the seizure data. The database data itself is stored separately, with access granted on request as described below.

Licence

The python tools are licenced under the GNU General Public Licence, V3.

The OpenSeizureDatabase data is licenced under a variation of the Creative Commons Share Alike licence, with an additional licence condition to publish a description of the system developed and results obtained using the data for the benefit of users of the OpenSeizureDetector project - see the LICENCE description for more details.

Data Users

The following people have access to the anonymised OpenSeizureDatabase data.

NameContactResearch AreaWeb Site
Graham Jonesgraham@openseizuredetector.org.ukDevelopment of Improved SeizureDetection Algorithms for OpenSeizureDetectorOpenSeizureDetector
Jamie Pordoypordjam@uwl.ac.ukPhD Research into Seizure DetectionIntSaV Group
Benjamin Mickler Research into seizure detection 

Researchers who would like access to the data should email osdb@openseizuredetector.org.uk explaining what they intend to use the data for, and confirming that they will comply with the requirements of the LICENCE for the data. They will then be given access to a repository containing the JSON files that make up the Open Seizure Database.

Installation Instructions

  • Create a folder in your home directory called "osd" and change directory into it.
  • Clone this repository (git clone https://github.com/OpenSeizureDetector/OpenSeizureDatabase.git)
  • change directory to ~/osd/OpenSeizureDatabase
  • create a python virtual environment with python -m venv ~/pyEnvs/osdb, and activate it with source ~/pyEnvs/osdb/bin/activate.
  • execute pip install -r requirements.txt (note the tensorflow requirement is not essential so can be removed if you do not intend to use tensorflow to train neural networks).
  • Create a folder in your home directory called "osd/osdb"
  • Copy the OSDB JSON text files into ~/osd/osdb

Test Installation

  • Go to ~osd/OpenSeizureDatabase/user_tools/dataSummariser
  • execute python ./summariseData.py --index. This should produce a file output/index.html which lists all the data in the database. Note that there will be missing image files because these are only generated when a full summary is created.
  • exeute python ./summariseData.py --event=5483. This should produce a folder, output/Event_5483_summary which contains an html file (index.html) and associated images to display a summary of the event, similar to the example output in the documentation folder.

Documentation

For details of the data structure and the software included in this repository, please refer to the Documentation. The documentation folder also contains an Example set of data for a seizure for reference.