ITT-8.1

AM-FM Texture Segmentation in Electron Microscopic Muscle Imaging
Marios S Pattichis (Washington State University), Constantinos S Pattichis, Maria Avraam (University of Cyprus), Alan C Bovik (The University of Texas at Austin), Kyriakos Kyriakou (Cyprus Institute of Neurology and Genetics)

We segment the structural units of electron microscope muscle images using a novel AM-FM image representation. This novel AM-FM approach is shown to be effective in describing sarcomeres and mitochondrial regions of the electron microscope muscle images.

ITT-8.2

Detection of Fetal ECG with IIR Adaptive Filtering and Genetic Algorithms
Amit Kam (Elect. & Computer Eng. Dept. Ben-Gurion Uni. Beer-Sheva, Israel), Arnon Cohen (Elect. & Computer Eng. Dept., Ben-Gurion Uni., Beer-Sheva, Israel)

The continuous monitoring of fetal heart condition during pregnancy and labor is of great clinical importance. The cardiac electrical activity of the fetus (FECG) may be recorded by means of surface abdominal electrodes. The signal is severely contaminated by the maternal cardiac signal (MECG). FECG enhancement is usually performed by FIR adaptive filtering. A new IIR FECG enhancement system is suggested and evaluated. In order to avoid convergence into local extremum, the system employs genetic algorithm (GA). Two architectures are considered. The first is a combination of adaptive filter and GA where the GA is recruited whenever the adaptive filter is suspected of reaching a local extremum. The second is an independent GA search. The hybrid IIR-GA algorithm was shown to be superior to the conventional FIR adaptive filtering.

ITT-8.3

Low Power Real-Time Programmable DSP Development Platform for Digital Hearing Aids
Trudy D Stetzler (Texas Instruments), Neeraj Magotra (University of New Mexico), Pedro R Gelabert (Texas Instruments), Preethi Kasthuri, Sridevi Bangalore (University of New Mexico)

This paper presents a new low power binaural wearable digital hearing aid platform based on the Texas Instruments TMS320C5000 fixed point digital signal processor. This platform is a real-time system capable of processing two input speech channels at a 32KHz sampling rate for each channel and driving a stereo headphone output. It provides for frequency shaping, noise suppression, multiband amplitude compression, and frequency dependent interaural time delay algorithms. Since the platform is a programmable solution capable of running at 1.8V for MIPS intensive research and 1V for actual hearing aid implementation, this platform will enable further research into improving the quality of life for the hearing impaired.

ITT-8.4

FRACTAL DIMENSION CHARACTERIZES SEIZURE ONSET IN EPILEPTIC PATIENTS
Rosana Esteller, George J. Vachtsevanos (School of Electrical and Computing Engineering, Georgia Institute of Technology, Atlanta, GA), Javier Echauz (Universidad de Puerto Rico, Mayaguez), Tom Henry, P Pennell, C Epstein, R Bakay, Christina Bowen, Brian Litt (Emory University School of Medicine, Atlanta, GA)

We present a quantitative method for identifying the onset of epileptic seizures in the intracranial electroencephalogram (IEEG), a process which is usually done by expert visual inspection, often with variable results. We performed a fractal dimension (FD) analysis on IEEG recordings obtained from implanted depth and strip electrodes in patients with refractory mesial temporal lobe epilepsy (MTLE) during evaluation for epilepsy surgery. Results demonstrate a reproducible and quantifiable pattern that clearly discriminates the ictal (seizure) period from the pre-ictal (pre-seizure) period. This technique provides an efficient method for IEEG complexity characterization, which may be implemented in real time. Additionally, large volumes of IEEG data can be analyzed through compact records of FD values, achieving data compression on the order of one hundred fold. This technique is promising as a computational tool for determination of electrographic seizure onset in clinical applications.

ITT-8.5

Development of Sound Source Components for a New Electrolarynx Speech Prosthesis
Kenneth M Houston (Draper Laboratory), Robert E Hillman, James B Kobler, Geoffrey S Meltzner (Massachusetts Eye and Ear Infirmary)

For many individuals who lose their voices due to laryngeal cancer or trauma, the only option for speech is to use an electrolarynx (EL), which is a battery-powered vibrator that is held to the throat. Current devices produce speech that is very machine-like in sound, with low levels of loudness and intelligibility, that also draws undesired attention to the user. A project at Draper Laboratory, the Mass. Eye and Ear Infirmary and MIT aims to develop a much improved EL called the Electrolarynx Communication System (ELCS), which is a DSP-based device consisting of sound source, control, and speech enhancement subsystems or modules. This paper introduces the ELCS and discusses developments to date in the sound source module. Specific topics include the design of a new linear EL transducer and investigations into glottal waveform synthesis which should result in a much more natural speech output.

ITT-8.6

Detection of Seizure Signals in Newborns
Boualem Boashash, Paul Barklem, Mark Keir (Signal Processing Research Centre - Queensland University of Technology)

This paper considers a system design for processing a multidimensional biomedical signal formed by EEG, ECG, EOG and motion recorded from a newborn, for the purpose of detection of epileptic seizures in newborns as an extension of the method reported in [1,8]. We describe the proposed design, and discuss how the signals will be analysed and fused to detect the occurrence of seizure. We also discuss the role of modelling in refining the signal processing unit.

ITT-8.7

Fast Detection of Masses in Digitized Mammograms
Ioanna Christoyianni (WCL, Electrical & Computer Engineering Dept.,University of Patras), Evangelos Dermatas (WCL, Electrical & Computer Engineering Dept., University of Patras), George Kokkinakis (WCL, Electrical & Computer Engineering Dept.,University of Patras)

A novel method for fast detection of regions of suspicion (ROS) that contain circumscribed lesions in mammograms is presented. The position and the size of ROS are first recognized with the aid of a Radial-Basis-Function neural network (RBFNN) by performing windowing analysis. Then a set of criteria is employed to these regions to make the final decision concerning the abnormal ones. Accelerated estimation of the high-order statistical features decreases the computational complexity 55 times in multiplication operations. The proposed method detects the exact location of the circumscribed lesions with accuracy of 72.7% (overlap between groundtruthed and detected regions greater than 50%) for mammograms containing masses, while the recognition rate for the normal ones reaches 77.7% in the MIAS database.

< ITT-7 ITT-9 >

Last Update: February 4, 1999 Ingo Höntsch