We report here on in vitro and in vivo experiments that are intended to explore the feasibility of photoacoustic spectroscopy as a tool for the noninvasive measurement of blood glucose. The in vivo results from oral glucose tests on eight subjects showed good correlation with clinical measurements but indicated that physiological factors and person-to-person variability are important. In vitro measurements showed that the sensitivity of the glucose measurement is unaffected by the presence of common blood analytes but that there can be substantial shifts in baseline values. The results indicate the need for spectroscopic data to develop algorithms for the detection of glucose in the presence of other analytes.

A time-resolved photoacoustic technique has been applied to the study of dissolved and dispersed absorbers in aqueous systems. The temporal pressure profiles generated from colloidal graphite and glucose solutions were measured, and it was found that the amplitude of the photoacoustic signal of both the glucose and the colloidal graphite solutions increase linearly with concentration and that acoustic signal time delay yields the acoustic velocity. The logarithm of the photoacoustic signal amplitude changes linearly with the time delay, with a slope that is proportional to the product of the acoustic velocity and the optical absorption that can thus be determined.

Glucose concentrations have been measured by the pulsed photoacoustic technique both in-vitro and in-vivo. The in- vitro results show that glucose concentrations in the physiological range (30 - 600mg/dl) can be measured with an accuracy of better than 0.1% with a correlation of R2 = 0.989. The role of other blood analytes in photoacoustic measurements have been explored, ¡η-vivo measurements on 8 subjects have been achieved with a standard deviation of between ± 1.5% and ± 12 % with an R2 = 0.959 over the glucose concentration range 80 - 640mg/dl.

No abstract