In vivo glucose detection using surface enhanced Raman spectroscopy

The work described in this dissertation details the steps taken towards an implantable surface-enhanced Raman spectroscopy (SERS) based glucose sensor. Extensive work had been done to validate the SERS approach for glucose sensing in vitro. Detailed herein is the path taken from in vitro testing to in vivo transdermal sensing.;Raman spectroscopy is an optical technique, therefore initial in vivo studies utilized a windowed dorsal skinfold chamber installed in a rat to allow optical access to an implanted sensor. Blood glucose levels were manipulated by injection of glucose and insulin through cannulated blood vessels. SER spectra were acquired simultaneously with readings from a conventional home blood glucometer. Using Partial Least Squares, a portion of the data set was used to create a calibration model from spectra correlated with the glucometer readings. The remaining data were used as independent point in a validation set. Good predictive values were shown, with relatively low error.;The successful initial in vivo results took place over a relatively short time frame, less than 6 hours. To succeed as an implanted glucose sensor, the lifetime of the device must be longer, on the order of five days or longer. Long term biocompatibility issues were investigated with miniaturized sensors implanted in rats for up to five days. Tissue surrounding the sensor was examined histologically for signs of inflammation and foreign body response and the recovered sensors were tested for viability ex vivo. Little difference between implant and control tissue was seen and the sensors remained SERS-active.;Installing a window is eminently impractical if the sensor is to be used as an implantable device. Studies were performed testing the spatially offset Raman spectroscopy (SORS) method to acquire spectral data through skin without the need for a dorsal skinfold chamber. Ex vivo and in vivo tests are reported with the AgFON sensor functionalized with the conventional DT/MH SAM as well as a benzenethiol (BZT) SAM, which produces a markedly larger SERS signal. Finally, the first in vivo transdermal SERS-SORS glucose measurements are reported, showing SORS as a valid method in use with the SERS sensor.