SUMMARY
UCLA researchers and physicians in the Department of Psychiatry have developed a software program to improve the analysis and clinical interpretation of human electroencephalogram (EEG) signals.
BACKGROUND
Electroencephalography (EEG) is a medical diagnostic test that uses electrodes to measure the spontaneous electrical activity of the brain. This electrical activity is representative of key activities and features of the underlying neurons that control brain activity. Therefore, EEGs are commonly used to diagnose and track various neurologic and psychiatric disorders. The ability of EEGs to track brain activity over time makes it an especially valuable tool to diagnose patients with epilepsy, sleep disorders and migraines. The non-invasive nature of EEG also makes it ideal for monitoring patients in comas or under anesthesia. While EEGs have been used for almost 100 years, the underlying dynamics that control the EEG have remained elusive. Recent advances in neuroscience have revealed that EEG signal changes as a function of time and therefore non-stationary and nonlinear. Traditional methods of EEG analysis assume that EEG signals are a stationary process, necessitating novel methods of analysis in order to most accurately portray the underlying dynamics of the EEG trace.
INNOVATION
Dr. Todd Zorick in the Department of Psychiatry at UCLA has created a software program for the analysis of human EEG signals. Unlike many current methods of EEG analysis, this program takes into account the non-stationary nature of EEG signals, allowing it more accurately portray the neuronal dynamics that underlie the EEG signal. This program also simplifies EEG analysis by transforming complex EEG waves over a period of time into a "DFA spectrum" in the form of a single curve. Application of this analysis to publically available human EEG traces demonstrates low patient-to-patient variation demonstrating the precise nature of the analysis. Additionally, this technology was able to distinguish different stages of sleep and identify an observed seizure. Therefore, this program has great potential to be used as an "EEG reader" to report and classify a patient's EEG results compared to a known database.
POTENTIAL APPLICATIONS
- Tool for identifying various sleep stages in sleep studies
- Tool for monitoring stages of general anesthesia during surgery
- Research tool to identify EEG abnormalities with neurologic/psychiatric disorders
- Tool for identifying a variety of psychiatric illnesses
- Tool for distinguishing seizure states
ADVANTAGES
- Increase the accuracy, consistency and level of EEG interpretation
- Widely applicable to aid the diagnosis of numerous neurologic and psychiatric disorders
- Better identify and distinguish patterns of neurologic abnormalities among patients with neurologic or psychiatric disorders
- Can quantify treatment results or assess clinical changes over time
STATE OF DEVELOPMENT
Several versions of the software have been written in a publicly available software package. This software has been successfully implemented on a small collection of human EEG signals and computer simulations. The program has successfully confirmed observed seizures and identified different stages of the sleep.