Photoacoustic Imaging in the Short-Wave Infrared: Advancements in the Detection of Deep-Tissue Constituents

Abstract

Tissue constituents such as lipids, water and collagen (LWC) are known to play crucial roles in numerous diseases and ailments. Changes of these constituents in biological tissue are associated with conditions ranging from cancer and Alzheimer’s disease to atherosclerosis and wound healing. It is henceforth necessary to have the capability to image and monitor the evolution of LWC in tissue with high resolution, specificity, and accuracy. Standard imaging modalities currently used to achieve this task mainly include MRI, CT and a variety of optical methods such as confocal microscopy and OCT. These modalities however exhibit shortcomings in terms of adequate spatial resolution, field of view, time acquisition and/or depth penetration which are needed to precisely track LWC at sub-superficial tissue depths. The goal of this research is to address these limitations by developing and demonstrating novel techniques for imaging LWC with enhanced resolution, depth, and specificity. Specifically, the work aims to advance photoacoustic imaging (PAI) to overcome the current challenges associated with standard imaging modalities. PAI is an emergent technique which allows for the high-resolution spatial mapping (~100 microns) of tissue absorption at depths > 5 mm, which is currently unmet in the broader field of medical imaging. While exhibiting the needed qualities for imaging constituents, PAI itself is not readily used to image LWC as the ability to generate contrast between the chromophores is limited due to low absorption coefficients within the typically utilized waveband of 680 - 1000 nm. In this dissertation, PAI is demonstrated in the short-wave infrared (SWIR, 1000 – 2000 nm) as a powerful and substantial technique to identify and image LWC in tissue. Techniques of PAI in the SWIR will be developed in both transmission and reflection-mode systems towards the application of probing constituents at depths > 5 mm. Beginning with a transmission-mode setup, higher-order spectral unmixing algorithms are constructed and shown to identify lipid/water content in excised murine brains, along with lipid signatures in ex vivo intramuscular tissue. Methods will be developed to implement these techniques into more clinically translatable reflection-mode systems, where the challenge of SWIR light delivery is directly addressed via the manufacture of a novel opto-acoustic coupling medium (patent pending). Rigorous material characterization and testing is conducted to obtain marks on optical and acoustic bandpass parameters which permit bi-directional sound and light delivery for reflection-mode systems. Techniques are further evolved with the in vivo size and location mapping of ischemic strokes in murine brains, where lipid coagulation is demonstrated as an endogenous biomarker for stroke infarct identification. Finally, system analysis will be conducted on a current in-lab PA apparatus, hypothesizing the implementation of these advanced techniques to expand its spectral and spatial resolution capabilities. In general, the progression of SWIR PAI as presented in this dissertation sets a benchmark for future studies aiming to probe LWC in tissue, ultimately leading clinicians to obtain much improved information about diseases related to such constituents.