Advanced Photothermal Optical Coherence Tomography (PT-OCT) for Quantification of Tissue Composition

Optical coherence tomography (OCT) is an imaging technique that forms 2D or 3D images of tissue structures with micron-level resolution. Today, OCT systems are widely used in medicine, especially in the fields of ophthalmology, interventional cardiology, oncology, and dermatology. Although OCT images provide insightful structural information of tissues, these images are not specific to the chemical composition of the tissue. Yet, chemical tissue composition is frequently relevant to the stage of a disease (e.g., atherosclerosis), leading to poor diagnostic performance of structural OCT images. Photo-thermal optical coherence tomography (PT-OCT) is a functional extension of OCT with the potential to overcome this shortcoming by overlaying the 3D structural images of OCT with depth-resolved light absorption information. Potentially, signal analysis of the light absorption maps can be used to obtain refined insight into the chemical composition of tissue. Such analysis, however, is complex because the underlying physics of PT-OCT is multifactorial. Aside from tissue chemical composition, the optical, thermal, and mechanical properties of tissue affect PT-OCT signals; system/instrumentation parameters also influence PT-OCT signals. As such, obtaining refined insight into tissue chemical composition requires in-depth research aimed at answering several key unknowns and questions about this technique. The goal of this dissertation is to generate in-depth knowledge on sample and system parameters affecting PT-OCT signals, to develop strategies for optimal detection of a molecule of interest (MOI) and potentially for its quantification, and to improve the imaging rate of the system. The following items are major outcomes of this dissertation: 1- Generated comprehensive theory that discovers relations between sample/tissue properties and experimental conditions and their multifactorial effects on PT-OCT signals. 2- Developed system and experimentation strategies for detection of multiple molecules of interest with high specificity. 3- Generated optimized machine learning-powered model, in light of the above two outcomes, for automated depth-resolved interpretation of tissue composition from PT-OCT images. 4- Increased the imaging rate of PT-OCT by orders of magnitude by introducing a new variant of PT-OCT based on pulsed photothermal excitation. 5- Developed algorithms for signal denoising and improving the quality of received signals and the contrast in images which in return enables faster PT-OCT imaging.