In an optically transparent environment, unlike for biologic tissues, the quantification of chiral substances such as glucose is based on the measurement of the amount of rotation of polarized light as it propagates a known distance through a fixed volume of the sample. This can be accomplished simply, using linear polarizers (e.g., polaroid sheet) to measure changes in detected polarized light intensity. This metric has proven particularly useful for the quantification of sugar and sucrose in the food and beverage industries. For decades now, there has been a concerted effort to implement this same metric for the noninvasive quantification of blood glucose levels (BGL), particularly as a diagnostic and monitoring tool for diabetes. The challenge is primarily the inability to exactly ascertain the actual pathlength traversed and the corresponding blood volume interrogated by the polarized light beam, due to tissue scattering.
The current paper addresses the aforementioned issues using a computational modeling approach. A genetic algorithm is used to systematically compute the most likely tissue and glucose contributions responsible for the experimentally detected back-reflected polarized light intensities measured from probing the scattering tissue sample. Using six—not entirely independent—different input states of polarization, a minimization process is conducted until the overall error between the simulated intensities using the assumed tissue and glucose contributions reaches a desired threshold. The authors demonstrate that glucose concentration obtained by this method correlates well with the actual experimental values utilized in a tissue-simulating phantom.
The authors plan to extend this work to the quantification of physiological levels of glucose using more representative biologic tissue samples. We look forward to seeing those results corroborated by in-vivo studies that use this method to noninvasively quantitate BGL.
You must log in to add comments.