3 Approximately 17 million people in the USA (6{\%} of the
4 population) and 140 million people worldwide (this number is
5 expected to rise to almost 300 million by the year 2025) suffer
6 from \textit{diabetes mellitus}. Currently, there a few dozens of
7 commercialised devices for detecting blood glucose levels [1].
8 However, most of them are invasive. The development of a
9 noninvasive method would considerably improve the quality of life
10 for diabetic patients, facilitate their compliance for glucose
11 monitoring, and reduce complications and mortality associated with
12 this disease. Noninvasive and continuous monitoring of glucose
13 concentration in blood and tissues is one of the most challenging
14 and exciting applications of optics in medicine. The major
15 difficulty in development and clinical application of optical
16 noninvasive blood glucose sensors is associated with very low
17 signal produced by glucose molecules. This results in low
18 sensitivity and specificity of glucose monitoring by optical
19 methods and needs a lot of efforts to overcome this difficulty.
21 A wide range of optical technologies have been designed in
22 attempts to develop robust noninvasive methods for glucose
23 sensing. The methods include infrared absorption, near-infrared
24 scattering, Raman, fluorescent, and thermal gradient
25 spectroscopies, as well as polarimetric, polarization
26 heterodyning, photonic crystal, optoacoustic, optothermal, and
27 optical coherence tomography (OCT) techniques [1-31].
29 For example, the polarimetric quantification of glucose is based
30 on the phenomenon of optical rotatory dispersion, whereby a chiral
31 molecule in an aqueous solution rotates the plane of linearly
32 polarized light passing through the solution. The angle of
33 rotation depends linearly on the concentration of the chiral
34 species, the pathlength through the sample, and the molecule
35 specific rotation. However, polarization sensitive optical
36 technique makes it difficult to measure \textit{in vivo} glucose
37 concentration in blood through the skin because of the strong
38 light scattering which causes light depolarization. For this
39 reason, the anterior chamber of the eye has been suggested as a
40 sight well suited for polarimetric measurements, since scattering
41 in the eye is generally very low compared to that in other
42 tissues, and a high correlation exists between the glucose in the
43 blood and in the aqueous humor. The high accuracy of anterior eye
44 chamber measurements is also due to the low concentration of
45 optically active aqueous proteins within the aqueous humor.
47 On the other hand, the concept of noninvasive blood glucose
48 sensing using the scattering properties of blood and tissues as an
49 alternative to spectral absorption and polarization methods for
50 monitoring of physiological glucose concentrations in diabetic
51 patients has been under intensive discussion for the last decade.
52 Many of the considered effects, such as changing of the size,
53 refractive index, packing, and aggregation of RBC under glucose
54 variation, are important for glucose monitoring in diabetic
55 patients. Indeed, at physiological concentrations of glucose,
56 ranging from 40 to 400 mg/dl, the role of some of the effects may
57 be modified, and some other effects, such as glucose penetration
58 inside the RBC and the followed hemoglobin glycation, may be
61 Noninvasive determination of glucose was attempted using light
62 scattering of skin tissue components measured by a
63 spatially-resolved diffuse reflectance or NIR fre\-quen\-cy-domain
64 reflectance techniques. Both approaches are based on change in
65 glucose concentration, which affects the refractive index mismatch
66 between the interstitial fluid and tissue fibers, and hence
67 reduces scattering coefficient. A glucose clamp experiment showed
68 that reduced scattering coefficient measured in the visible range
69 qualitatively tracked changes in blood glucose concentration for
70 the volunteer with diabetes studied.
