Blood glucose monitoring is a primary tool for the care of diabetic patients. At present there is no non-invasive monitoring technique of blood glucose concentration that is widely accepted in the medical industry. Some promising success has been shown by the RIT ETA Lab research group that an antenna's resonant frequency can track, in real time, changes in glucose concentration. However, these changes in antenna response would be unique to an individual depending on various parameters such as the person's metabolism, body mass index, chemical profile etc. Therefore, amongst the several issues that still need to be addressed in the development of a noninvasive methodology; an important one is the calibration unique to each individual patient. The objective of the present work is to establish a calibration by developing an equivalent circuit model to represent the antenna input impedance with changing glucose levels. Using the calibration, the glucose level is estimated from the measurement of antenna impedance and resonant frequency at a later date. Measurements were made on a diabetic and non-diabetic subject with a blood glucose meter and with a planar dipole antenna connected to a network analyzer. The network analyzer was automated to make input impedance measurements in real-time using Labview in 15 second intervals and calculates the resonant frequency of the antenna. The resonant frequency of the antenna and the reference glucose values using a blood glucose meter were measured. The resonant frequency of the antenna is plotted in real-time for instant analysis and trends. A linear estimation for the glucose based on the resonant frequency of the antenna was compared to reference measurements taken by a blood glucose meter. 59% estimates were with 20% of the reference measurements for the diabetic patient. An equivalent circuit model, developed to estimate glucose concentration, was found to have a shunt resistance, series capacitance, and shunt inductance. The component values for this circuit are generated from the experimental results of the diabetic patient, unique to the subject. The component values are iteratively varied until they reflect the impedance of the input impedance. Because the input impedance is changing over time, the component values change over time. The calibration model uses fitted equations to the component values from the experimental result that relates the component values to a glucose concentration. This unique equation for the diabetic patient can be used to estimate the glucose level when measuring the input impedance of the antenna at a later date. The experiment was repeated on the diabetic subject and the unique calibration procedure was used to estimate the blood glucose concentration. The estimates represented by the Clarke's Error Grid compared favorably to the measurements made by the blood glucose meter.
Library of Congress Subject Headings
Blood sugar monitoring--Technological innovations; Medical instruments and apparatus--Calibration
Department, Program, or Center
Electrical Engineering (KGCOE)
Sidley, Matthew, "Calibration for real-time non-invasive blood glucose monitoring" (2013). Thesis. Rochester Institute of Technology. Accessed from
RIT – Main Campus