For results that have...

But don't include...

Abstract

Click to add/remove this article to your list of 'My Favorites' NLDS Electrical Spectroscopy Technique For Continuous Non-invasive Determination Of Blood Glucose - Initial Evaluation In Non-Diabetic Human Subjects

Year: 2007

Abstract Number: 0425-P

Authors: EOIN BAMBURY, FRAN KEATING, ADRIAN MCHUGH, JOHN O[apos]DEA, Galway, Ireland

Results: Non-linear Dielectric Spectroscopy (NLDS) is an electrical technique proposed to non-invasively determine blood glucose concentration in a continuous manner. The technique relies on transducing reflected harmonics, which are generated in a glucose concentration dependent fashion, due to the conformational change of membrane-bound glucose receptor proteins when they are stimulated by a low frequency, low voltage, electrical field. The results of a preliminary evaluation of the technique on human subjects are presented herein.
A linear array of four carbon loaded polyester electrodes was applied to the shaven abdomen of each subject. The four-electrode technique uses two electrodes to stimulate the cells, and two electrodes to measure the harmonic response. The electrodes were connected to a PC-based data acquisition system (Model PCI-6024E Data Acquisition Card driven using Labview V7.0 Software, National Instruments, Austin, TX).
Each subject, after 10 hours fasting, was given a carbohydrate supplement drink (Polycal, Nutricia Clinical Care, England). Reference blood glucose measurements were taken with a hand held glucose meter (Optium Plus, Medisense, England). Individually optimized algorithms, which estimate blood glucose based on NLDS measurements, were retrospectively derived for each subject.
Table 1 presents the first published test results using the technique applied to five human subjects. Data presented includes mean and median absolute relative difference (ARD) between measured and algorithm-derived glucose measurements for each subject. All data-pairs fell within the A (79.6%) and B (20.4%) zones of the Clark Error Grid.

Case# MeasurementsGlucose Range
(mg/dL)
Mean ARD (%)Median ARD (%)
11797-1687.96.6
22676-16612.812.4
32592-24610.89.9
42193-18116.514.1
52471-16917.111.5

Table 1 : MARD for five non-diabetic subjects tested
Further research is being undertaken to determine suitable prospective calibration algorithms, electrode material choice and geometry,and to test the technique over a wider glucose range.