Multispectral polarimetric sensor for glucose monitoring utilizing a digital closed-loop control system

Abstract

Treatment of many medical disorders, such as diabetes hics. and other hormonal or metabolic imbalances, requires accurate blood analysis. Conventionally, blood is analyzed by withdrawing a sample from the body of the subject. One of the major disadvantages of conventional methods is the invasive nature of the tests that raise the risk of patient infection and discomfort. The polarimetric approach is currently being researched to determine glucose levels in the body non-invasively. Polarized light experiences a change in its plane of rotation that is proportional to the concentration of the sample when passed through an optically choral sample such as glucose. A multi-wavelength, multi-beam polarimetric device has been designed and implemented. The potential site for in-vivo measurements is the aqueous humor of the eye, which shows glucose concentrations proportional to the concentrations in the blood. The purpose of this dual wavelength device is to provide the means for overcoming two important problems with in-vivo glucose monitoring, namely, motion artifacts and the presence of other optically active substances at the test site. The device was tested in-vitro for accuracy and sensitivity. It predicted glucose to within an average standard deviation of 9.88mg/d1 for 670nm wavelength and 7.61mg/dl for the 820nm wavelength for the hyperglycemic ranges of 0-600mg/dl of glucose doped water. Improved results were obtained for glucose predictions in the hypoglycemic range of 0-100mg/dl of glucose. Glucose was predicted to within a standard deviation of 9.57mg/d1 for the 67013m wavelength and 5.35mg/dl for the 820nm wavelength. The system was also able to extract glucose information to within an average standard deviation of 24.41mg/dl from a solution of albumin and glucose, albumin being the other choral substance used in this study. Motion artifact studies indicated a clear trend of the two wavelengths in tracking each other, which potentially could be used to allow for glucose measurements in the presence of motion artifacts. The novelty of the system was its simultaneous dual wavelength approach and a fast feedback control system implemented in Labview. Overall, the results are satisfactory and show considerable being developed as an important approach to glucose potential in this technique sensing.