Journal article
Calibration Strategy for Noninvasive Glucose Measurements Based on Near-Infrared Spectroscopy: in Vitro Simulations
Analytical letters, Vol.59(11), pp.2019-2036
2026
DOI: 10.1080/00032719.2025.2578671
Abstract
A calibration protocol for noninvasive blood glucose measurements based on near-infrared spectroscopy was investigated with in vitro samples from a synthetic biological matrix. The methodology used a combination of direct orthogonal signal correction (DOSC) spectral preprocessing, followed by application of hybrid linear analysis to implement the calibration model. The goal of the study was to devise a calibration procedure that would require minimal effort on the part of a diabetic for both an initial calibration and subsequent updating of the calibration over time. A set of calibration data was collected to compute the initial model, and subsequent model updates were based on the collection of four samples with constant glucose concentrations. These procedures were designed to mimic noninvasive measurements with one data collection session involving an oral glucose tolerance test for the initial calibration, followed by model updates based on data collected when the glucose level was constant (e.g., at the beginning of a day after fasting overnight). Using these procedures, the standard error of prediction (SEP) for glucose was maintained between 0.13 and 0.22 mM over the 1–20 mM range for 281 days after calibration. Incorporation of the updating protocol improved the SEP values on average by 90% when compared to the results obtained with the original calibration model. Use of DOSC in conjunction with updating improved the SEP by 29% when compared to the use of updating alone. The proposed methodology provides a pathway forward for the practical implementation of noninvasive glucose measurements.
Details
- Title: Subtitle
- Calibration Strategy for Noninvasive Glucose Measurements Based on Near-Infrared Spectroscopy: in Vitro Simulations
- Creators
- Austin J. Gessell - University of IowaGary W. Small - University of Iowa
- Resource Type
- Journal article
- Publication Details
- Analytical letters, Vol.59(11), pp.2019-2036
- DOI
- 10.1080/00032719.2025.2578671
- ISSN
- 0003-2719
- eISSN
- 1532-236X
- Publisher
- TAYLOR & FRANCIS INC
- Grant note
- Samsung Advanced Institute of Technology
Funding for this work was provided by the Samsung Advanced Institute of Technology.
- Language
- English
- Electronic publication date
- 10/24/2025
- Date published
- 2026
- Academic Unit
- Chemistry
- Record Identifier
- 9985019029002771
Metrics
8 Record Views