EEGNet: Collaborative Software for EEG Data Review and Annotation


EEGnet is a novel Web-based JavaScript platform for enabling collaborative multi-site research on large scale EEG (brain wave) data, and for individual users to access, annotate, and manipulate EEG files from any web-capable computer. Users may scroll by 10-s epochs through the entire sample, adjust channel gain and montage settings, add or adjust digital filters, and annotate regions of the EEG recording (either in single channels or on all channels). Using a web-based interface, it allows researchers at different physical sites to view and interact with large collections of EEG data.  EEGnet can be used for a variety of research and training purposes; for example, experts can annotate EEG data that can then be used to automatically test performance of automated methods for seizure or spike detection or localization, or novices can use EEGnet as a tutorial system to better improve their skills at EEG analysis. EEGnet has many other useful features as well (e.g., viewing EEG in multiple montages, several different methods of annotation, simple user and trial management, ability to display time-locked video and more).


EEGnet displays standard 10-20 montage EEG data in a web browser. The EEG data is stored on the EEGnet server and transferred via the internet to the web browser second-by-second as the user scrolls through the data. Usually, the display of the standard EEG data requires the use of expensive commercial software which is installed on a personal computer. The EEGnet software makes it much easier for physicians to review EEG remotely from any web-connected PC. The uses for the EEGnet system include (1) clinical EEG telemedicine projects, (2) collection of expert opinion from physicians at remote locations on the waveforms present in EEG, and (3) education of physicians on EEG interpretation. The third generation of the EEGnet software (EEGnet v3), allows for greater control of 2D graphics rendering and allow easier implementation of new features due to increased software modularity.


EEGnet is the only web-based software system in the world which can display standard 10-20 montage EEG in the way it is displayed on standard commercial clinical EEG workstation used by neurologists. The EEGnet system currently includes these features:

1.       Montaging and the ability to import custom montages

2.       Global and individual channel gain control

3.       Local and flexible control of digital high pass, low pass, and notch filtering coded in Javascript (HTML5)

4.       No EEG or video signal data stored locally, because all EEG data streamed live to the web browser second-by-second

5.       Individual user login, individual user to-do lists, and the server logs annotation actions of each individual user in real-time

6.       No limitations on the total number of users that can be logged into the system at one time, except for the computational power of the EEGnet server array (which is scalable).

7.       Multiple EEG annotation functions, including the ability to highlight EEG on a signal channel and the ability to mark epochs of EEG including all channels, on any montage

8.       The system can read EEGLab and EDF file format.

9.       Ability to view EEG and annotate EEG recordings of any length.

10.       Ability to view video which has been simultaneously recorded with EEG (although the common proprietary EEG formats from the OEMs have not been deciphered yet).

11.       Collection of expert opinion annotations to the EEGnet server in real time with no storage of expert opinion at the local PC. This enables progress of experts in annotating EEG datasets to be tracked in real time.


Overview: Inventors noticed one of the primary impediments in the field of routine scalp electroencephalogram (rsEEG) research was a lack of an efficient method for collecting the opinion of large numbers of EEG experts on specific features within EEG recording. In order to do this type of review, prior to EEGnet, EEG recordings had to be mailed out on CDs to EEG experts, and the experts had to annotate their opinions on paper or by adding text annotations to the EEG recordings themselves using commercial EEG review software which had to be installed on a PC locally. The only other web based iEEG portal, is also an EEG viewing and annotation system, but it is customized to work with intracranial EEG, not 10-20 montage scalp EEG, and lacks many of the features of EEGnet for viewing standard 10-20 montage scalp EEG data.


EEGnet can be used to collect expert opinion on the two most common types of EEG recording performed in the clinical environment: routine EEG (rEEG) and inpatient long-term monitoring (LTM) EEG. The rEEG is the most common clinical neurophysiology procedure. It consists of a 20–30 min recording from approximately 20 scalp electrodes. Over one million outpatient rEEGs and over 50,000 inpatient rEEGs are performed in the United States every year. The most important role of rEEG is to detect evidence of epilepsy, in the form of epileptiform transients (ETs), also known as spike or sharp wave discharges. Due to the wide variety of morphologies of ETs and their similarity to artifacts and waves that are part of the normal background activity, the task of ET detection is difficult and mistakes are frequently made by interpreting physicians. Despite the importance of rEEG, little has changed in the typical way that rEEG is recorded and interpreted over the last 25 years. The development of reliable computerized detection of ETs in the EEG could assist physicians in interpreting rEEGs. No reliable systems for automated rEEG interpretation are currently available.


The fastest growing type of EEG recording is LTM EEG, which is usually performed in the intensive care unit (ICU) setting. LTM EEG studies are usually between six hours and one week in duration. The purpose of LTM EEG recording is to detect the presence of seizure activity, but LTM EEG monitoring is also performed monitor other patterns in the EEG such as depth of IV sedation and rhythmic ETs. No reliable method of automated interpretation of LTM EEG is currently available.


Applications: Viewing and collaboration surrounding 10-20 montage scalp EEG data for training and research purposes.

Advantages:  EEGnet allows much easier large-scale distributed review, annotation, and analysis of EEG data than was previously possible, enabling easier validation of new algorithms, technologies, and devices.

Key Words: EEG, research, software, analysis, annotation, trial management, data


Publications: Halford, Jonathan J., et al. "Interictal Epileptiform Discharge Detection in EEG in Different Practice Settings." Journal of Clinical Neurophysiology 35.5 (2018): 375-380.


Bagheri, Elham, et al. "Interictal epileptiform discharge characteristics underlying expert interrater agreement." Clinical Neurophysiology 128.10 (2017): 1994-2005.


Foreman, Brandon, et al. "Generalized periodic discharges and ‘triphasic waves’: A blinded evaluation of inter-rater agreement and clinical significance.“ Clinical Neurophysiology (2015).


Halford, J. J., et al. "Inter-rater agreement on identification of electrographic seizures and periodic discharges in ICU EEG recordings." Clinical Neurophysiology (2014).


Halford, Jonathan J., et al. "Standardized database development for EEG epileptiform transient detection: EEGnet scoring system and machine learning analysis." Journal of neuroscience methods 212.2 (2013): 308-316.


Halford, Jonathan J., et al. "Web-based collection of expert opinion on routine scalp EEG: software development and interrater reliability." Journal of Clinical Neurophysiology 28.2 (2011): 178-184.


Inventors: Brian Dean PhD, Jonathan Halford MD

Patent Status:       Copyright

MUSC-FRD Technology ID: P1479


Patent Information:
For Information, Contact:
Kaitlyn Crobar
Technology Transfer Associate
MUSC Foundation for Research Development
Jonathan Halford
Brian Dean
© 2022. All Rights Reserved. Powered by Inteum