New developments in biomedical signaling processing have produced new tools for the understanding of biological/biomedical phenomena. Using these methods we have found a direct experimental link between speech production, the hearing derived psychoacoustic mel-scale, and the basilar membrane. Our results provide an explanation about how communication is possible, in spite of speech being produced from physiologically different sources, for example, adult vocal tract versus a child's vocal-tract. Based on these results a direct aim of the proposed research is the improvement of cochlear implants, which currently suffer serious degradation in intelligibility in the presence of noise. We also apply these new methods to balance problems and propose new time-frequency methods to understand the response of individuals to visual stimulus. It has been shown that the response produces a non- stationary signal and hence time-frequency methods are essential. Understanding balance and having markers for detection of early onset is particularly important for the aging population, which is one of the groups we will study. Also, we propose new medical imaging methods for image enhancement, texture analysis, and image registration. The proposed research aims at detecting and classifying texture, which is often an indicator of diseased tissue, and aims to developing methods that will enhance images by fusion images obtained from different sources or modalities, which is common in diagnostic medical imaging.
Showing the most recent 10 out of 202 publications
