Faculty Advisor or Committee Member

Stevan Kun, Advisor

Faculty Advisor or Committee Member

Sergey N. Makarov, Committee Member

Faculty Advisor or Committee Member

Raymond M. Dunn, Committee Member


Robert A. Peura




"Ischemia is a condition of decreased tissue viability caused by a lack of perfusion, which prevents the delivery of oxygen and nutrients to biological tissue. Ischemia plays a major role in many clinical disorders, yet there are limited means by which tissue viability can be assessed. The long-term objective of this research is to develop a non-invasive or non-contact instrument for quantifying human tissue ischemia. Skeletal muscle ischemia is evaluated at this stage because skeletal muscle is easily accessible, its ischemia represents a clinical problem, and it can endure short periods of ischemia without suffering permanent injury. The ischemia monitor designed for this study is based on impedance spectroscopy, the measurement of tissue impedance at various frequencies. This study had three major goals. The first goal was to improve upon the design of the ischemia monitor to achieve optimal system performance in a clinical environment. Major considerations included electrode sterility, instrument mobility, and electrosurgical unit interference. The second goal was to collect both impedance and pH data from human subjects undergoing tourniquet surgeries, which induce skeletal muscle ischemia and result in changes of the tissue's pH and impedance. The average in recorded pH during ischemia was 0.0053 pH units/minute and the average change in Ro was -0.1481 Ohms/minute. The third goal was to develop a relationship between parameters of tissue impedance and pH utilizing neural networks. This goal was accomplished in three stages. First, the optimal neural network type for classifying impedance data and pH values was determined. Based on these results, the backpropagation neural network was utilized for all subsequent work. Then, the input parameters of the neural network were optimized using previously collected data. The number of inputs to the previously developed neural network were reduced by 35% (13/20) with a maximum of a 3% reduction in neural network performance. Finally, the neural network was trained and tested using human impedance and pH data. The network was able to correctly estimate tissue pH values with an average error of 0.0440 pH units. Through the course of this research the ischemia monitor based on impedance spectroscopy was improved, a methodology for the use of the instrument in the operating room was developed, and a preliminary relationship between parameters of impedance spectra and pH was established. The results of this research indicate the feasibility of the instrument to monitor both pH and impedance in a clinical setting. Additionally, it was demonstrated that impedance data collected non-invasively could be used to estimate the pH and level of ischemia in human skeletal muscle."


Worcester Polytechnic Institute

Degree Name



Biomedical Engineering

Project Type


Date Accepted





impedance spectroscopy, non-invasive instrumentation, ischemia, Impedance spectroscopy, Ischemia, Muscles