Date of Graduation

2004

Document Type

Thesis

Degree Type

MS

Committee Chair

Ronald L. Klein

Abstract

The research presented here was carried out at the National Institute for Occupational Safety and Health (NIOSH), a division of the Centers for Disease Control and Prevention (CDC) in Morgantown, West Virginia. At NIOSH, there is a strong need for modes of analyzing a wealth of biomedical data. In this particular case, that data was comprised of numerous muscleinduced force-time series, derived from two groups of experimental subjects. One group was comprised of rats exposed to vibration similar to that experienced by human workers who work with certain hand-tools or equipment. The other group was a control group, not exposed to the vibration. Humans exposed to vibration on a consistent basis are known to develop a condition known as hand-arm vibration syndrome (HAVS). Thus, there was a need for a non-invasive method to differentiate between the two groups, while providing a means of quantifying the effects of vibration. Two signal algorithms were developed using MATLAB (The Mathworks, Natick, MA). The first algorithm performs a complex analysis of a given time-series, first computing several time-domain quantities and then converting to the frequency domain to perform Fourier-based spectral analysis on the signal. The program computes a number of time-series measures, such as maximum value, mean value, and standard deviation. Welch’s averaged modified periodogram method of spectral estimation is then used to compute the power spectral density, or spectrum, of the signal. Once in the frequency domain, a number of intricate analyses are carried out on the signal. Measures computed in the frequency domain include total power, and the amplitude and frequency location of the five largest spectral peaks. The spectrum is also divided into a number of overlapping frequency bins. The power inside each frequency bin is computed, as well as the maximum value inside each bin and the frequency at which that value occurs. The second signal processing algorithm computes a relatively novel measure known as approximate entropy. Approximate entropy (ApEn) is a measure of the amount of regularity (or, conversely, randomness) in a signal. Results showed both major and subtle differences between groups. There were notable differences between the vibration and non-vibration groups with respect to average force, total power, peak amplitude, peak frequency location, ApEn, and other measures. Some significant differences were noted across time, as well. There were effects of force and time. Some results indicate possible changes in motor unit recruitment patterns and frequencies for both groups. If so, changes in motor unit recruitment for the vibration group were different from those of the non-vibration group. Overall, the techniques of spectral analysis and regularity quantification presented here are shown to be effective methods for analysis of biomedical data. Specifically, these techniques are shown to be useful in the analysis of musculoskeletal force-time signals.

Download

Share

COinS