Share Email Print

Proceedings Paper

Factor analysis for comparing NIR reflection and transmittance for noninvasive glucose measurement
Author(s): Kye Jin Jeon; Sangjoon Hahn; In Duk Hwang; Gilwon Yoon
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

For noninvasive measurement of bio-fluid substances in human body based on optical spectroscopy, optical measurement system is one of the most important parts. We studied glucose specificities by analyzing the factors in the partial least squares regression models for the two cases of reflectance and transmittance measurements. Glucose -intralipid solutions were used as the samples whose scatterer's concentrations were varied. We used intralipid concentrations of 4%, 4.08% and 4.16% in the solution and these values were comparable to tissue scattering. Temperature was maintained at 30°C during measurement. Factor analysis for reflectance data didn't show glucose absorption feature and the factors were very noisy particularly in the combination band. It is speculated that light does not have enough information of glucose since the pathlength in reflectance is very short. On the other hand, the factors obtained from the PLS analysis of transmittance revealed glucose signatures. We suggest that transmittance measurement is preferred for in vivo glucose monitoring than reflectance measurement.

Paper Details

Date Published: 1 July 2004
PDF: 8 pages
Proc. SPIE 5318, Advanced Biomedical and Clinical Diagnostic Systems II, (1 July 2004); doi: 10.1117/12.529728
Show Author Affiliations
Kye Jin Jeon, Samsung Advanced Institute of Technology (South Korea)
Sangjoon Hahn, Samsung Advanced Institute of Technology (South Korea)
In Duk Hwang, Samsung Advanced Institute of Technology (South Korea)
Gilwon Yoon, Seoul National Univ. of Technology (South Korea)

Published in SPIE Proceedings Vol. 5318:
Advanced Biomedical and Clinical Diagnostic Systems II
Gerald E. Cohn; Warren S. Grundfest M.D.; David A. Benaron M.D.; Tuan Vo-Dinh, Editor(s)

© SPIE. Terms of Use
Back to Top