Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group
Journal Home About Issues in Progress Current Issue All Issues Feature Issues

Noninvasive, optical detection of diabetes: model studies with porcine skin

E.L. Hull, M.N. Ediger, A.H.T. Unione, E.K. Deemer, M.L. Stroman, and J.W Baynes

Open Access Open Access

Abstract

An in vitro study was performed to evaluate noninvasive spectroscopic measurement of advanced glycation endproducts (AGEs) in skin collagen. A porcine dermis preparation was incubated in solutions simulating normal and hyperglycemic conditions. The AGEs kinetics of increase were determined by HPLC and GC/MS assays, and compared to near-infrared (NIR) and ultraviolet/visible fluorescence skin spectra. Multivariate analysis indicated that, although NIR did not discriminate between collagen samples exposed to different glucose concentrations, fluorescence changes were readily detected and correlated strongly with skin concentration of AGEs. These results suggest that measurement of skin AGEs by fluorescence spectroscopy may be useful for detection and diagnosis of type II diabetes.

©2004 Optical Society of America

Full Article  |  PDF Article
More Like This
Optoacoustic detection of tissue glycation

Ara Ghazaryan, Murad Omar, George J. Tserevelakis, and Vasilis Ntziachristos
Biomed. Opt. Express 6(9) 3149-3156 (2015)

Multiphoton spectral microscopy for imaging and quantification of tissue glycation

Jo-Ya Tseng, Ara A. Ghazaryan, Wen Lo, Yang-Fang Chen, Vladimir Hovhannisyan, Shean-Jen Chen, Hsin-Yuan Tan, and Chen-Yuan Dong
Biomed. Opt. Express 2(2) 218-230 (2011)

Skin autofluorescence for the risk assessment of chronic complications in diabetes: a broad excitation range is sufficient

M. Koetsier, H. L. Lutgers, A. J. Smit, T. P. Links, R. de Vries, R. O. B. Gans, G. Rakhorst, and R. Graaff
Opt. Express 17(2) 509-519 (2009)

View More...

Cited By

Optica participates in Crossref's Cited-By Linking service. Citing articles from Optica Publishing Group journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (9)
Fig. 1.
Fig. 1. Accumulation of CML (left) and pentosidine (right) during incubation.

Download Full Size | PDF

Fig. 2.
Fig. 2. Mean intra-subject variance of corrected fluorescence (left) and correlation between subject age and corrected fluorescence (right) as function of correction factors kx and km.

Download Full Size | PDF

Fig. 3.
Fig. 3. Schematic representation of a single cross-validation iteration. Validation data (red box) consisted of all spectra from a given specimen collected on a single study day (in this case, the fourth normoglycemic specimen on the third study day). Calibration data (blue boxes) consisted of spectra from all specimens except the validation specimen (i.e., the vertical column of gray boxes) obtained on all days except the validation study day (i.e., the horizontal row of gray boxes). Calibration data were mean-centered according to their incubation medium in order to eliminate medium-specific biases that could unfairly inform the PLS model. The mean spectrum of the calibration data from the appropriate incubation medium was subtracted from the validation data prior to generation of PLS AGE estimates.

Download Full Size | PDF

Fig. 4
Fig. 4 Average NIR reflectance spectra for all specimens in each incubation medium. Each trace depicts the average spectrum of all specimens in a given incubation medium on one of the 10 days of spectral data acquisition during the 5-week experiment. Vertical axes are identical for all three panels.

Download Full Size | PDF

Fig. 5
Fig. 5 Average fluorescence excitation spectra (λ x = 315-385 nm, λ m = 400 nm) for all specimens in each incubation medium.

Download Full Size | PDF

Fig. 6
Fig. 6 Cross-Validated Standard Errors of Prediction for CML and Pentosidine using both NIR data sets (transmission, reflectance; dashed lines), the three independent fluorescence data sets (M1, X1, X2; thinner solid lines), and the three fluorescence data sets appended together into a single large spectrum (thick solid line). ● = statistically significant result, □= optimum number of PLS model factors as computed by Akaike’s Information Criterion [37].

Download Full Size | PDF

Fig. 7
Fig. 7 PLS predictions from analysis of the appended fluorescence spectral data set vs. wet-chemical assay results for CML (left panel) and pentosidine (right panel).

Download Full Size | PDF

Fig. 8
Fig. 8 Control CVSEP curves for CML and Pentosidine using both NIR data sets (transmission, reflectance; dashed lines), the three independent fluorescence data sets (M1, X1, X2; thinner solid lines), and the three fluorescence data sets appended together into a single large spectrum (thicker solid line). For the control cases, the reference values for the hyperglycemic and normoglycemic specimens were switched. ● = statistically significant result.

Download Full Size | PDF

Fig. 9.
Fig. 9. Left panel: CML concentrations from wet chemical assays of punch biopsy specimens taken from Type 1 diabetic patients and nondiabetic controls (reproduced from Ref. 16). Right panel: Receiver-Operator Characteristic (ROC) curves calculated by cross-validated quadratic discriminants analysis of the data in the left panel under assumptions noninvasive measurement errors of zero (blue line), 0.5 mmol/mol lysine (green line), and 1.0 mmol/mol lysine (red line). The measurement error encountered in the in vitro experiment was 1.24 mmol/mol lysine.

Download Full Size | PDF

Equations (3)

Equations on this page are rendered with MathJax. Learn more.

f xm = F xm R x k x R m k m
AGE = b ̂ · s val ,
CVSEP = ∑ i = 1 N pred e i 2 / ( N pred − 3 ) ,
Select as filters
Select Topics Cancel
Top
       
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.
Privacy | Terms of Use