Chair: Nitish V. Thakor, Johns Hopkins School of Medicine (USA)
Carlos E. Davila, Southern Methodist University
Richard Srebro, University of Texas Southwestern Medical Center
Masoud Azmoodeh, Southern Methodist University (USA)
Ibrahim Ghaleb, Southern Methodist University (USA)
The eigenfilter is an FIR filter that maximizes signal-to-noise ratio (SNR). It typically consists of the eigenvector associated with the maximum eigenvalue of the data covariance matrix. Alternately, the eigenfilter may incorporate a linear combination of the dominant covariance matrix eigenvectors. Expressions for the eigenfilter SNR gain are derived. An algorithm for adaptive eigenfiltering is then described which has a computational complexity of $O(Md^2)$ where $M$ is the eigenfilter length and $d$ is the signal covariance matrix rank. The algorithm is demonstrated via simulations to out-perform a well-known subspace averaging algorithm having similar computational complexity. The eigenfiltering algorithm is then used to obtain estimates of the single trial steady-state visual evoked potential.
Vijay Parsa, University of New Brunswick (CANADA)
Philip Parker, University of New Brunswick (CANADA)
Robert Scott, University of New Brunswick (CANADA)
Somatosensory Evoked Potentials (SEPs) are extremely useful in peripheral nerve monitoring and in the diagnosis of various neuromuscular disorders. However, surface measurements of these potentials often result in imperceptible SEP waveforms due to the very poor Signal-to-Noise Ratio (SNR). Adaptive noise cancelling is an attractive technique which can be used to improve this poor SNR. One of the important factors that effect the performance of an Adaptive Noise Canceller (ANC) is the presence of SEP components in the reference channel of the ANC. In this paper we propose a novel Multicha- nnel Crosstalk Resistant Adaptive Noise Canceller (MCRANC) for offsetting the problems caused by the SEP crosstalk. The performance of this MCRANC is evaluated analytically and through simulations.
Bogdan R. Kosanovi, University of Pittsburgh (USA)
Luis F. Chaparro, University of Pittsburgh (USA)
Robert J. Sclabassi, University of Pittsburgh (USA)
In this paper, we present a method that is a generalization of hidden Markov modeling for the situations where elementary events cannot be clearly defined. A family of fuzzy sets, induced on a temporal universe, is used to model the dynamic trajectory of a physical system as a collection of hidden processes that coexist at the same time, but to different degrees. An algorithm based on unsupervised pattern recognition that estimates the prototypes and activities of the hidden processes is presented. The performance of the method is illustrated using experimental data obtained from electroencephalographic (EEG) signals recorded during sleep.
David L. Sherman, Johns Hopkins University School of Medicine (USA)
Yien Che Tsai, Johns Hopkins University School of Medicine (USA)
Lisa Ann Rossell, Johns Hopkins University School of Medicine (USA)
Marek A. Mirski, Johns Hopkins University School of Medicine (USA)
Nitish V. Thakor, Johns Hopkins University School of Medicine (USA)
Time series analysis applications of eigenstructure algorithms focus on temporal frequency estimation . We show that the ESPRIT algorithm can also be applied to simple phase delays for sinusoids. We show that a time delay data model can be rendered in the ESPRIT matrix pencil structure. The PRO-ESPRIT formulation can be then utilized to solve for phase delays among sinusoids. An application area for this algorithm is the estimation of short time delays for low frequency sinusoids comprising EEG (electroencephalographic) recordings derived from different neural sites during epileptic seizure activity.
John C. Mosher, Los Alamos National Laboratory
Richard M. Leahy, University of Southern California
Paul S. Lewis, Los Alamos National Laboratory (USA)
The most widely used models for electroencephalography (EEG) and magnetoencephalography (MEG) assumes a quasi-static approximation of Maxwell's equations and a piecewise homogeneous conductor model. Both models contain an incremental field element that linearly relates an incremental source element (current dipole) to the field or voltage at a distant point. The field element can be partitioned into the product of a vector dependent on sensor characteristics and a matrix kernel dependent only on head modeling assumptions. Proper characterization of this element is crucial to the inverse problem. We present here the matrix kernels for the general boundary element model (BEM) and for MEG spherical models. We show how these kernels are interchanged in a linear algebraic framework that includes sensor specifics such as orientation and gradiometer configuration. We then describe how this kernel is applied to gain or transfer matrices used in multiple dipole and source imaging models.
Xuan Kong, Northern Illinois University
Vaibhava Goel, Johns Hopkins University (USA)
Nitish Thakor, Johns Hopkins University (USA)
Accurate detection and characterization of changes in EEG signal is crucial for clinical assessment of the neurological system condition. Several distance measures are tested and evaluated for their effectiveness of detecting injury- related changes in EEG. Itakura distance is found to be a very efficient means to characterize changes in EEG for both signaling injury and predicting recovery. The efficiency of the Itakura distance measure is further established through a comparison study of spectral distance measure and KL information.