Data acquisition unit for low-noise, continuous glucose monitoring

As the number of people with diabetes continues to increase, research efforts improving glucose testing methods and devices are under way to improve outcomes and quality of life for diabetic patients. This dissertation describes the design and testing of a Data Acquisition Unit (DAU) providing low noise photocurrent spectra for use in a continuous glucose monitoring system. The goal of this research is to improve the signal to noise ratio (SNR) of photocurrent measurements to increase glucose concentration measurement accuracy. The glucose monitoring system consists of a portable monitoring device and base station. The monitoring device measures near infrared (IR) absorption spectra from interstitial fluid obtained by microdialysis or ultrafiltration probe and transmits the spectra to a base station via USB or a ZigBee radio link. The base station utilizes chemometric calibration methods to calculate glucose concentration from the photocurrent spectra. Future efforts envisage credit card-sized monitoring devices. The glucose monitor system measures the optical absorbance spectrum of an interstitial fluid (ISF) sample pumped through a fluid chamber inside a glucose sensor. Infrared LEDs in the glucose sensor illuminate the ISF sample with IR light covering the 2.2 to 2.4 micron wavelength region where glucose has unique features in its absorption spectrum. Light that passes through the sample propagates through a linearly variable bandpass filter and impinges on a photodiode array. The center frequency of the variable filter is graded along its length such that the filter and photodiode array form a spectrometer. The data acquisition unit (DAU) conditions and samples photocurrent from each photodiode channel and sends the resulting photocurrent spectra to the Main Controller Unit (MCU). The MCU filters photocurrent samples providing low noise photocurrent spectra to a base station via USB or Zigbee radio link. The glucose monitoring system limit of detection (LOD) from a single glucose sensor wavelength is 5.8 mM with a system bandwidth of 0.00108 Hz. Further analysis utilizing multivariate calibration methods such as the net analyte signal method promise to reduce the glucose monitoring system LOD approaching a clinically useful level of approximately 2 mM.