QCL and FT-IR Spectroscopic Imaging to Identify Biomarkers for Transplant Outcome and Cancer

The large disjoint between engineering and medicine has increasingly become a major focus of the current generation of researchers. Efforts are now are being made to bridge the gap between these fields. While engineering has made many advances, this technology is not always readily translated into clinical use. One such technique that has high potential for clinical use but hasn’t been readily implemented is Infrared (IR) spectroscopic imaging. The main focus of this proposal is translating IR imaging, a powerful chemical imaging technique, in to the clinical setting as an adjunct to current techniques. IR imaging is an emerging modality that can allow for insight into the biochemistry of tissues/cells in an entirely label-free and non-perturbing fashion. IR imaging is based on the principle that chemical bonds absorb mid-infrared light quantitatively and can allow for the simultaneous measurement of biomolecules in tissues and cells including proteins, lipids, carbohydrates, DNA and RNA. The past several years has seen large advances in the field of IR imaging with dramatic increases in the speed of data acquisition to allow for imaging of tissues within minutes, increased spatial resolutions of up to 1 x 1 microns and advances in computational approaches for data-mining these feature-rich spectroscopic data sets. With advances in IR imaging, there have been new technology appearing in the field of IR spectroscopic imaging. The most recent push of new technology has led to the use of quantum cascade lasers (QCL) as a new type of source of IR and thus overcoming some of the limitations of conventional Fourier Transform Infrared (FT-IR) spectroscopy. But unlike QCL systems, FT-IR cannot perform real time imaging and other features that would make IR spectroscopy more clinically feasible. Leveraging unique aspects of QCL systems is an important step towards clinical translation. In this study, we leverage bioinformatics and biomedical imaging techniques (such as feature extraction, feature selection and machine learning algorithms) in effort to automate and streamline assessment of tissue. We apply these techniques in various setting as cancer and organ transplantation. Specifically, we have applied these techniques to probe allograft status in a label-free manner for states such as diabetic nephropathy (transplant and primary kidneys), progressive interstitial fibrosis (transplant kidneys), and anti-body mediated rejection (transplant heart). We have also applied developed methodologies for increasing understanding of liver cancer progression for diabetic patients and discriminating between the difficult diagnoses of thyroid carcinoma subtypes.