An optical modality capable of quantitative, label-free, high-speed and high-resolution imaging across spatiotemporal scales coupled with sophisticated software for image reconstruction and quantitative analyses would be of great utility to scientists and engineers in the medical and life sciences fields. Currently, a combination of optical imaging techniques and software packages are needed to address the list of capabilities described previously. Optical coherence tomography is an optical imaging technique based on low coherence interferometry capable of measuring light backscattered from the sample at micrometer-level resolutions over millimeter-level penetration depths in biological tissue. Phase-sensitive extensions of OCT enable the functional assessment of biological tissue samples as well as the structural examination of samples down to the single-cell level. This dissertation describes the development and application of high-speed real-time multi-functional spectral-domain OCT (MF-SD-OCT) for structural and functional examination of biological samples across spatiotemporal scales. A discussion of the development of a GPU-accelerated high-speed MF-SD-OCT imaging system accompanied by demonstrations of the performance enhancements due to the GPU are presented initially. Next, the development of MF-SD-OCT-based quantitative methods for the structural and functional assessment and characterization and classification of biological tissue samples is discussed. The utility of these methods is demonstrated through structural, functional and optical characterization and classification of peripheral nerve and muscle tissue. The dissertation concludes with a discussion of the improvements made to spectral-domain optical coherence phase microscopy (SD-OCPM) to enable dynamic live cell imaging and the application of dynamic live cell SD-OCPM for morphological visualization of cheek epithelial cells and examination of functionally stimulated morphological changes in neurons. The work described in this dissertation demonstrates the versatility of OCT imaging technology to perform quantitative, label-free, high-speed and high-resolution imaging across spatiotemporal scales, enabling the examination of native dynamic tissue and cellular physiology.

Cerebral edema is a condition characterized by a net increase in brain water content. Typically water content is tightly regulated in the brain. However this balance can be disrupted by a number of conditions including traumatic brain injury (TBI), tumor and ischemia and lead to brain swelling, raised intracranial pressure (ICP) and secondary damage that greatly increases the morbidity and mortality associated with these conditions. Intracranial pressure is the most common method for detecting and monitoring cerebral edema in vivo. However, this technique is only capable of providing a global assessment of brain water content and is unable to detect cerebral edema early in its development. The goal of the work presented here is to develop optical coherence tomography (OCT) as a tool for detecting cerebral edema in vivo. OCT is an optical imaging modality capable of producing cross-sectional images of biological samples with micrometer resolution. This work begins by studying an in vivo global edema model with OCT and analyzing the change in the depth-resolved attenuation coefficient. Results demonstrated that the attenuation coefficient decreases due to edema. Additionally, decreased blood flow caused by severe brain swelling was also observed. Next, a focal model of edema is explored in a TBI mouse model. Scattering changes correlated with bleeding and edema formation were observed around the impact site. Lastly, OCT system modifications are described for creating a differential absorption-OCT system designed to increase the OCT signal sensitivity to water in particular, by using the absorption properties of water to determine local water content in the tissue. Preliminary results are presented. Overall the results of this work demonstrate that OCT is sensitive to changes in scattering and absorption caused by cerebral edema and highlight the potential of OCT for in vivo cerebral edema detection and monitoring in a spatially resolved manner.

We demonstrate a computationally-efficient method for optical coherence elastography (OCE) based on fringe washout method for a spectral-domain OCT (SD-OCT) system. By sending short pulses of mechanical perturbation with ultrasound or shock wave during the image acquisition of alternating depth profiles, we can extract cross-sectional mechanical assessment of tissue in real-time. This was achieved through a simple comparison of the intensity in adjacent depth profiles acquired during the states of perturbation and non-perturbation in order to quantify the degree of induced fringe washout. Although the results indicate that our OCE technique based on the fringe washout effect is sensitive enough to detect mechanical property changes in biological samples, there is some loss of sensitivity in comparison to previous techniques in order to achieve computationally efficiency and minimum modification in both hardware and software in the OCT system. The tissue phantom study was carried with various agar density samples to characterize our OCE technique. Young’s modulus measurements were achieved with the atomic force microscopy (AFM) to correlate to our OCE assessment. Knee cartilage samples of monosodium iodoacetate (MIA) rat models were utilized to replicate cartilage damage of a human model. Our proposed OCE technique along with intensity and AFM measurements were applied to the MIA models to assess the damage. The results from both the phantom study and MIA model study demonstrated the strong capability to assess the changes in mechanical properties of the OCE technique. The correlation between the OCE measurements and the Young’s modulus values demonstrated in the OCE data that the stiffer material had less magnitude of fringe washout effect. This result is attributed to the fringe washout effect caused by axial motion that the displacement of the scatterers in the stiffer samples in response to the external perturbation induces less fringe washout effect.

Optical coherence tomography (OCT) is capable of high-resolution cross-sectional imaging of retinal tissue and is typically paired with near-infrared (NIR) wavelengths due to its ability to penetrate deeper into tissue compared to visible wavelengths. This imaging modality has helped ophthalmologists diagnose and learn more about the structural changes found in retinal diseases. To help further the effectiveness and versatility of this imaging modality, broadband visible wavelengths were incorporated into the system to enable detection of spectroscopic signals. Wavelengths centered at 580 nm will provide valuable molecular information that was not provided with just NIR wavelengths. Here we demonstrated the use of visible wavelengths and introduced a novel spectroscopic analysis technique in a visible OCT system, which was then characterized and tested.

Optical coherence tomography (OCT) is a novel and promising imaging technique that has been widely applied in clinical practice and research. Currently advanced spectral-domain OCT is capable for fast generating three dimensional (3D) volumetric image and extracting multi-biological information that has potential uses in disease study. Based on a laboratory-built multi-functional spectral-domain OCT system with center wavelength of 1310 nm, the research goal in this thesis is to solve problems about proper optical characterization method of volumes of in vivo biological sample taken at different time points which requires appropriate 3D volume registration and alignment. In this thesis, system optimization is introduced in order to improve OCT signal-to-noise ratio (SNR) and image quality for better image registration performance. Details of a computational-efficient 3D volume alignment based on cross-correlation image registration is discussed. Optical characterization of post-traumatic epilepsy (PTE) is described as an example to demonstrate the feasibility of the OCT system and to evaluate the image post-processing method in actual neural disease research.

ABSTRACT OF THE DISSERTATION

Quantification of Changes of Intensity and Attenuation Coefficient During Neural Activity Using Optical Coherence Tomography

by

Md. Monirul Hasan

Doctor of Philosophy, Graduate Program in Bioengineering

University of California, Riverside, March 2018

Dr. B. Hyle Park, Chairperson

The nervous system, which receives information through the electrical signal to coordinate and disseminate information about the body and its environment, is an assembly of neurons. Current methods of detection of neural activity are based on electrophysiology, which requires direct or near direct contact, or fluorescence-based techniques, which require the introduction of reporter molecules, raising a concern of toxicity. Optical coherence tomography(OCT) detects neural activity by taking advantage of intrinsic structural changes that accompany neural activity without any exogenous agents. The phase-resolved OCT detected 10-30 nm swelling of axon during neural activity from the functionally stimulated compound eye of the horseshoe crab(limulus) by utilizing the phase-resolved measurement. There was a decrease of backscattered intensity during propagation of action potential due to change of scattering as well. Further study on in vitro murine brain slice showed the changes of the phase (or thickness change) as well as a decrease of back-reflected intensity during propagation of action potential. Above studies with pr-OCT were performed without scanning because scanning introduced phase noise, and the desired signal was buried under noise for the signal to noise ratio of 25~30 dB. However, spectral domain OCT was used to capture 3D volumetric intensity images during the neural activity in in vitro seizure mouse model after a period of baseline data. The average intensity of the hippocampal brain slice decreased by 10~18% from baseline during neural activity. Although intensity can separate neural activity from non-neural activity with 97.27% specificity and 97.60% sensitivity, it varies from experiment to experiment. Attenuation coefficient(µ), which quantifies how strongly a sample scatters and absorbs light is a constant for a sample, can solve the problem. During neural activities, there was a 10~15% decrease of µ. For both intensity and µ based method, there was a spatial correlation between optical changes and gold standard electrophysiology. Finally, quantified OCT intensity detected a localized action potential from CA1 of the hippocampal brain slice during ripple activation and -1% intensity threshold separated activity from non-activity.

Optical coherence tomography (OCT) is a high resolution, minimally invasive imaging technique, which can produce depth-resolved cross-sectional images. OCT was used to detect changes in the optical properties of cortical tissue during the induction of global (pentylenetetrazol) and focal (4-aminopyridine) seizures. To utilize the spatiotemporal resolution provided by OCT, we developed an analytical method, creating fOCT volumes, to display statically significant changes in the cortical tissue’s attenuation coefficients while maintaining spatial specificity. We were able to visualize depth-resolved changes in attenuation during seizures as well as track focal seizure propagation with high spatiotemporal resolution. The results from this study demonstrate the potential utility of OCT as a minimally-invasive tool for studying seizure onset and propagation with high spatiotemporal resolution.

Molecular dynamics (MD) is a simulation technique that has been utilized to analyze biomolecular systems. Classical MD simulations and constant pH MD simulations have been applied to observe the behavior of components involved in two biomolecular systems of interest, the complement system and CRISPR-Cas9. The complement system is a defense mechanism part of the innate immune system. Complement associated diseases, such as paroxysmal nocturnal hemoglobinuria and atypical hemolytic uremic syndrome, are treated with complement inhibitors. Monoclonal antibody, eculizumab, is used as treatment for these diseases and functions as an inhibitor of complement component 5 (C5). A next generation version of eculizumab has also been developed known as ravulizumab, resulting from mutations within the heavy chain of eculizumab. MD simulations elucidated key residues involved in intermolecular interactions between complement inhibitors, eculizumab and ravulizumab, and C5. CRISPR (clustered regularly interspaced short palindromic repeats)-Cas9 is part of the bacterial adaptive immune system that functions as a genome editing tool. The HNH domain of CRISPR-Cas9 is involved in DNA cleavage. Through classical MD simulations, residues near the catalytic center of the HNH domain were observed. Mutations were applied to several residues within the HNH domain. The wild type structure was compared to different mutated structures to analyze the effect of the mutations. Distances between residues and RMSD were calculated. Constant pH MD simulations determined pKa values for histidine for the wild type and mutated structures. Taken together, our simulations clarified mechanisms and function of the complement and CRISPR-Cas systems, helping fundamental understanding and engineering.