Spatially Resolved Diffuse Reflectance Spectroscopy of Intestinal Tissue Using Concentric Silicon Thin Film Photodiode Arrays

Abstract

Patient outcomes for colon and esophageal cancers depend on early detection and surveillance of the progression of dysplasia (abnormal cellular growth indicative of pre-cancer) and malignancy (cancer) in the epithelium of these organs. The present standard of care for detection and surveillance of these diseases is to inspect the surface of the colon and esophagus searching for high-risk tissue using white light endoscopy (WLE), which provides imaging of the organ surface. While WLE is effective for the detection of some high-risk tissues such as colon polyps, other types of high-risk tissue, such as flat dysplasia, may not be visually observable to the physician by WLE. The limited efficacy of WLE results in reduced adenoma detection rates, unnecessary resection of healthy tissues, incomplete resection of high-risk tissues, and necessitates random sampling of esophageal or colon regions known to be at elevated risk for flat dysplasia, such as in cases of Barrett’s esophagus or inflammatory bowel disease. Patient outcomes for colon and esophageal cancer can be improved if the capabilities of WLE can be augmented to readily detect the microstructural and biochemical characteristics of these diseases in tissues during endoscopy. Diffuse reflectance spectroscopy (DRS) is a label free optical technique that can be used to sense the morphological and biochemical changes that occur when normal colon or esophageal tissue becomes pre-cancerous or cancerous, and DRS has the potential to be used to augment the performance of traditional WLE. However, the reliance of conventional DRS systems on optical fibers for reflectance collection limits system performance, limits scalability, and constrains the design options. Custom thin film silicon (Si) photodetectors (PDs) offer a high performance, scalable, and highly customizable alternative to optical fibers for reflectance photon collection in tissue DRS applications. Si PDs improve performance over optical fibers since they have superior numerical apertures (NAs) and collect light from a larger fraction of the probe surface; they are highly scalable due to Si processing technology; they can be geometrically optimized for specific applications; and they can be implemented on a variety of substrates, including flexible or conformal, and can be implemented as an attachment to an endoscope. And arrays of Si PDs can be used perform spatially resolved DRS (SRDRS), which adds depth sensitivity to DRS by simultaneously measuring the tissue reflectance at multiple illumination-collection distances. This dissertation describes the design, fabrication, optoelectronic characterization, phantom studies with reduced scattering coefficient (μ_s^' (λ)) and absorption coefficient (μ_a (λ)) extraction, and preliminary clinical testing on ex-vivo human colon tissue of the first Si thin film multi-pixel SRDRS array optimized for a specific tissue measurement application. The illumination-collection geometry of the Si PD based SRDRS sensor array described herein was designed using Monte Carlo simulation to maximize performance when characterizing human colon tissue. The design was then implemented using 10 μm thick crystalline Si PD arrays heterogeneously integrated onto a transparent substrate and packaged for phantom and clinical ex-vivo human tissue testing. The performance of the fabricated sensors was initially evaluated on synthetic liquid tissue phantoms, with comparisons of the experimentally measured diffuse reflectance from these phantoms compared to Monte Carlo simulated diffuse reflectance, yielding a mean discrepancy of 8.4% over the 450 to 650 nm wavelength range. To enable extraction of tissue optical properties from measured tissue diffuse reflectance, two inverse DRS methods were developed which relied on look-up tables of Monte Carlo simulated diffuse reflectance. One method extracts spectrally constrained values of μ_s^' (λ) and μ_a (λ) from measurements of diffuse reflectance independently for each PD in the SRDRS sensor. A second model uses the radial dependence of the PDs in the SRDRS sensor to both spatially and spectrally constrain the extracted values of μ_s^' (λ) and μ_a (λ). The average percent error between expected (known) and extracted μ_s^' (λ) and μ_a (λ) values of eight liquid phantoms used to evaluate the inverse models was 13.25% and 31.42%, respectively, for the first inverse method, and 6.38% and 8.38%, respectively, for the second, for diffuse reflectance measurements spanning the 450 nm to 750 nm wavelength range. These results indicate that use of three PDs in the SRDRS sensor can reduce the error for tissue optical property extraction. Further assessment of the Si PD array performance and tissue compatibility was conducted on ex-vivo porcine esophageal tissue, and an experimental set-up to enable clinical measurements of resected human colon tissue samples was developed in collaboration with physicians in the Duke University Medical Center: Dr. N. Lynn Ferguson, MD, Assistant Professor of Pathology in the Duke University Department of Pathology, Dr. Katherine Garman, MD, Associate Professor of Medicine in the Duke University Department of Medicine, and Dr. Deborah Fisher, MD, Associate Professor of Medicine in the Duke University Department of Medicine.SRDRS measurements on excised normal and tumorous human colon tissue were performed with a prototype thin film Si PD based SRDRS probe, and the absorption and reduced scattering optical properties of the tissue were extracted. The extracted optical properties are comparable to previously reported values optical properties for similar tissue and indicate that the prototype system is able to extract similar tissue information as other more expensive and less scalable tissue spectroscopic systems. The thin film Si PD based SRDRS probe and experimental work presented herein contributes to the field of Si PD based tissue DRS by being the first demonstration of a thin-film Si PD based DRS or SRDRS system in direct contact with human and animal tissues. The development, experimental validation, and demonstrated application to analysis of SRDRS measurements of human colon tissue, of a method for tissue optical property extraction described herein is the first inverse method to demonstrate an improvement of optical property extraction in a Si PD based system by utilizing the spatial dependence of reflectance, compared to a single PD based extraction method. The novel inverse model extraction method presented herein, which utilizes multiple Monte Carlo populated LUTs of diffuse reflectance to spatially and spectrally constrain extracted optical properties, is a contribution to the field of tissue optical property extraction from SRDRS measurements. Finally, the report of μ_s^' (λ) and μ_a (λ) properties for unfrozen ex-vivo human colon tissue adds to the sparsely reported data of this type in the literature, contributing to the general field of biophotonics.