Adaptive Fourier Modeling Of Evoked Potentials: Application To Cerebral Focal Ischemia
Paul W.F. Poon, Nitish V. Thakor · 2005
This paper presents a method for estimating evoked potentials (EPs) in a critical care environment where unpredictable transients may occur. Timely detection of these transients may prevent neurological injury. We have developed (1) a dynamic Fourier series model (FSM) for EPs and (2) an adaptive estimator based on the least mean square (LMS) and exponentially weighted least squares (EWLS) algorithms. Through computer simulations, we show that this technique is effective in both removing high EEG noise and in tracking rapid transients. Cerebral focal ischemia experiments were also conducted and confirmed the effectiveness of this technique. More interestingly, we also observed dispersion of EP spectral power which strongly suggested the importance of frequency content in diagnosis.