Optimized Detection of Hypoglycemic Glucose Ranges in Human Serum by Raman Spectroscopy with 532 nm Laser Excitation

Raman scattering-based biomedical detection has usually been proposed with near-infrared laser sources. However, a low-cost CMOS imager's quantum efficiency is optimum around green wavelength, and their sensitivity substantially decreases in near-infrared wavelengths. Additionally, since Raman scattering intensity is proportional to lambda(-4) , where lambda is the laser wavelength, the increase of wavelength directly results in less sensitive measurement. These facts contribute to limiting the transfer of detection methodologies based on Raman spectroscopy to portable and low-cost point-of-care medical devices. Therefore, here we propose 532 nm green laser-induced Raman spectroscopy for low human serum glucose level detection. However, in 532 nm Raman spectroscopy of carotenoid containing biological systems, such as human serum, resonance Raman occurs, and total carotenoids resonance bands dominate the spectra. To demonstrate serum glucose detection on concentration levels typical in severe hypoglycemic ranges, this study optimizes laser focal depth, laser excitation duration, and laser power to extend the sensitivity by exploiting the glucose Raman shift peak at 1125 +/- 7.5 cm(-1). By applying experimentally tuned parameters, our findings suggest sensitive detection of serum glucose in the range of 0-10 mmol/l with 1.2 mmol/l theoretical limit of detection (LOD) by using spontaneous (non-enhanced) Raman spectroscopy.