Advancing solid-state near-infrared spectroscopy for clinical measurements of glucose and urea



Document Type


Date of Degree

Fall 2012

Access Restrictions

Access restricted to UI faculty, staff and students.

Degree Name

PhD (Doctor of Philosophy)

Degree In


First Advisor

Arnold, Mark A

First Committee Member

Gillan, Edward G

Second Committee Member

Gloer, James B

Third Committee Member

Leddy, Johna

Fourth Committee Member

Margulis, Claudio J

Fifth Committee Member

Jessop, Julie


A solid-state based near-infrared (NIR) spectrometer is described and evaluated for the clinical measurements of glucose and urea. The development of a novel solid-state spectrometer advances noninvasive sensing technology by providing instrumentation capable of simple, robust, and reliable analytical measurements. The resulting solid-state systems promise the established features of NIR spectroscopy, including nondestructive, noninvasive, and reagent-less measurements. These features make solid-state NIR sensing attractive for bedside or near patient clinical applications associated with the treatment of diabetes and renal failure.

A solid-state based optical micro-sensor is described and a preliminary evaluation for continuous glucose monitoring is detailed. Partial least-squares (PLS) and net analyte signal (NAS) calibration models are developed based on in vitro data to demonstrate the analytical features of selectivity, sensitivity and accuracy for the measurements of glucose, urea, and lactate in a phosphate buffer matrix. To further extend the analysis, the measurement of glucose in an animal model is demonstrated.

A digital micro-mirror array device (DMD) based solid-state spectrometer is developed for selective analytical measurements and potentially for multivariate optical computing. A novel optical configuration enabled characterization of the resolution of the DMD spectrometer by coupling the DMD output with an FTIR spectrometer. Other measurements demonstrate an improvement in the measurement signal-to-noise ratio (SNR) by 3.2-fold over earlier generations of the DMD spectrometer.

A selectivity analysis is performed for glucose in the presence of the following glycolysis intermediates: glucose 6-phosphate, fructose 6-phosphate, fructose 1,6-bisphosphate, phosphoenol pyruvate, and pyruvate. PLS calibration methods realize selectivity for glucose in this complex matrix and NAS analysis is used to understand the basis of selectivity under these demanding conditions.

Feasibility of an acousto-opto tunable filter (AOTF) based solid-state spectrometer is established for on-line monitoring of urea in spent dialysate during hemodialysis treatments. Six pure components and 60 mixtures of urea, glucose, lactate, creatinine, alanine, and glutamine, which represent the spent dialysate, are used to define analytical performance. Selectivity of the calibration models is characterized by a pure component selectivity analysis (PCSA). A prospective study demonstrates the long term potential of this system as a robust and stable calibration method for the accurate measurements of urea.


xiv, 183 pages


Includes bibliographical references (pages 173-183).


Copyright 2012 Joo Young Choi