Mid-Infrared photoacoustic detection of glucose in human skin: towards non-invasive diagnostics
J Kottmann, JM Rey, MW Sigrist - Sensors, 2016 - mdpi.com
J Kottmann, JM Rey, MW Sigrist
Sensors, 2016•mdpi.comDiabetes mellitus is a widespread metabolic disease without cure. Great efforts are being
made to develop a non-invasive monitoring of the blood glucose level. Various attempts
have been made, including a number of non-optical approaches as well as optical
techniques involving visible, near-and mid-infrared light. However, no true breakthrough has
been achieved so far, ie, there is no fully non-invasive monitoring device available. Here we
present a new study based on mid-infrared spectroscopy and photoacoustic detection. We …
made to develop a non-invasive monitoring of the blood glucose level. Various attempts
have been made, including a number of non-optical approaches as well as optical
techniques involving visible, near-and mid-infrared light. However, no true breakthrough has
been achieved so far, ie, there is no fully non-invasive monitoring device available. Here we
present a new study based on mid-infrared spectroscopy and photoacoustic detection. We …
Diabetes mellitus is a widespread metabolic disease without cure. Great efforts are being made to develop a non-invasive monitoring of the blood glucose level. Various attempts have been made, including a number of non-optical approaches as well as optical techniques involving visible, near- and mid-infrared light. However, no true breakthrough has been achieved so far, i.e., there is no fully non-invasive monitoring device available. Here we present a new study based on mid-infrared spectroscopy and photoacoustic detection. We employ two setups, one with a fiber-coupled photoacoustic (PA) cell and a tunable quantum cascade laser (QCL), and a second setup with two QCLs at different wavelengths combined with PA detection. In both cases, the PA cells are in direct skin contact. The performance is tested with an oral glucose tolerance test. While the first setup often gives reasonable qualitative agreement with ordinary invasive blood glucose measurements, the dual-wavelength approach yields a considerably improved stability and an uncertainty of only ±30 mg/dL of the blood glucose concentration level at a confidence level of 90%. This result is achieved without advanced data treatment such as principal component analysis involving extended wavelength ranges.
MDPI