Accounting for Timing Drift and Variability in Contemporary Electroencepholography (EEG) Systems
Abstract
Electroencepholography (EEG) data acquisition technology is rapidly advancing. Because it is now more feasible to use in operational environments, its use in Army-relevant applications has also increased. However, integration of EEG with other systems requires the synchronization of multiple data modalities and identification of potential sources of error. Here, we assess the degree of stimulus timing drift and encoding jitter associated with several contemporary EEG systems. The measured amount of drift and jitter is then applied to several waveforms, including simulated, real evoked potentials, and data from a phantom head model for the sake of assessing the functional impact of each variable. Results show a wide range of drift across systems and particularly high timing variance in one; simulations show the potential for a dramatic distortion of signals if unaccounted for within those systems with more than 0.01% drift relative to an external clock. Additional simulations show a severe impact of encoding variance. We discuss a correction algorithm, with evidence that it not only alleviates distortion caused by drift but can also improve the results yielded from signal-averaging analyses in the case of high-variance systems.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 2012
- Accession Number
- ADA561715
Entities
People
- W. D. Hairston
Organizations
- United States Army Research Laboratory