Date of Award
Master of Science in Electrical Engineering - (M.S.)
Electrical and Computer Engineering
Stanley S. Reisman
Thomas W. Findley
The basic concepts and fundamentals of the wavelet signal representation were examined. The orthonormal wavelet was selected for this project after being compared to various types of wavelets. The orthonormal wavelet was chosen due to the equal time and frequency resolution exhibited in the wavelet coefficients. Programs were written in Matlab to implement the orthonormal wavelet in developing wavelet coefficients for a given signal. The programs include the conditions for an orthonormal wavelet in and which produce the wavelet filters g(n) and h(n). The wavelet filters were then incorporated into another program that applied Mallat's multiresolution algorithm for a given signal. The resulting wavelet coefficients were obtained and interpreted. The orthonormal wavelet was applied to various types of biomedical signals. The wavelet transform was applied to motor evoked potentials (MEPs) created cortical magnetic stimulation. The wavelet was also applied to evoked potentials (EPs) and to various types of EKG signals. The wavelet representation exposed new ways of observing biomedical signals by bringing out details and structures not present in the original waveforms.
Hauck, Karl O., "Application of the wavelet transform to biomedical signals" (1994). Theses. 1228.