Format

Send to

Choose Destination
Source Code Biol Med. 2013 May 21;8(1):12. doi: 10.1186/1751-0473-8-12.

EEGgui: a program used to detect electroencephalogram anomalies after traumatic brain injury.

Author information

1
Department of Neurology, 1095 NW 14th Terrace, Lois Pope LIFE Center, Miami, FL 33136, USA. TSick@miami.edu.

Abstract

BACKGROUND:

Identifying and quantifying pathological changes in brain electrical activity is important for investigations of brain injury and neurological disease. An example is the development of epilepsy, a secondary consequence of traumatic brain injury. While certain epileptiform events can be identified visually from electroencephalographic (EEG) or electrocorticographic (ECoG) records, quantification of these pathological events has proved to be more difficult. In this study we developed MATLAB-based software that would assist detection of pathological brain electrical activity following traumatic brain injury (TBI) and present our MATLAB code used for the analysis of the ECoG.

METHODS:

Software was developed using MATLAB(™) and features of the open access EEGLAB. EEGgui is a graphical user interface in the MATLAB programming platform that allows scientists who are not proficient in computer programming to perform a number of elaborate analyses on ECoG signals. The different analyses include Power Spectral Density (PSD), Short Time Fourier analysis and Spectral Entropy (SE). ECoG records used for demonstration of this software were derived from rats that had undergone traumatic brain injury one year earlier.

RESULTS:

The software provided in this report provides a graphical user interface for displaying ECoG activity and calculating normalized power density using fast fourier transform of the major brain wave frequencies (Delta, Theta, Alpha, Beta1, Beta2 and Gamma). The software further detects events in which power density for these frequency bands exceeds normal ECoG by more than 4 standard deviations. We found that epileptic events could be identified and distinguished from a variety of ECoG phenomena associated with normal changes in behavior. We further found that analysis of spectral entropy was less effective in distinguishing epileptic from normal changes in ECoG activity.

CONCLUSION:

The software presented here was a successful modification of EEGLAB in the Matlab environment that allows detection of epileptiform ECoG signals in animals after TBI. The code allows import of large EEG or ECoG data records as standard text files and uses fast fourier transform as a basis for detection of abnormal events. The software can also be used to monitor injury-induced changes in spectral entropy if required. We hope that the software will be useful for other investigators in the field of traumatic brain injury and will stimulate future advances of quantitative analysis of brain electrical activity after neurological injury or disease.

Supplemental Content

Full text links

Icon for BioMed Central Icon for PubMed Central
Loading ...
Support Center