Alzheimer's disease (AD) is a progressive neurodegenerative disease that affects millions of individuals in the United States alone. Common symptoms of the disease are forgetfulness and memory loss. However, these AD symptoms typically appear later in life despite potential early and hidden biological changes in the brain. This preclinical stage can began years before the onset of the typical symptoms of AD marking the need for earlier detection methods for developing therapies to slow symptom progression. Here, I have developed an initial susceptibility weighted imaging (SWI) method, a subset of magnetic resonance imaging, for the purpose of longitudinal study of AD.

Non-invasive visualization of the trigeminal nerve through advanced MR sequences and methods like tractography is important for studying anatomical and microstructural changes due to pathology like trigeminal neuralgia (TN), facial dystonia, multiple sclerosis, and for surgical pre-planning. The use of specific anatomical markers from CT, MPRAGE and cranial nerve imaging (CRANI) sequences, enabled successful tractography of patient-specific trajectory of the frontal, nasociliary, infraorbital, and mandibular nerve branches extending beyond the cisternal brain stem region and leading to the face. Performance of MPRAGE sequence together with the advanced T2-weighted CRANI sequence with and without a gadolinium contrast agent, was studied to characterize identification efficiency in smaller nerve structures in the extremities. A large FOV nerve visualization exam inclusive of the anatomy of all trigeminal nerve distal branches can be obtained within an acquisition time of 20 minutes using pre-contrast CRANI and MPRAGE. Post-processing with MPR and MIP images improved nerve visualization.Transcranial electrical stimulation techniques (TES) have been used for the treatment of multiple neurodegenerative diseases. These techniques involve placing electrodes on the scalp with multiple peripheral branches of the trigeminal nerve crossing directly under that may be stimulated. This was studied through hybrid computational realistic axon models. These models also facilitated studying the effects of electrode drift during experiments on the recruitment of peripheral nerves. An optimal point of lowest threshold was found while displacing the nerve horizontally i.e., the activation thresholds of both myelinated and unmyelinated axons increased when the electrodes were displaced medially and decreased to a certain extend when the electrodes were displaced laterally, after which further lateral displacement led to increase of thresholds. Inclusion of unmyelinated axons in the modeling provided the capability of finding maximum stimulation amplitude below which side effects like pain sensation may be avoided. In the case of F3 – F4 electrode montage the maximum amplitude was 2.39 mA and in case of RS – LS montage the maximum amplitude was 2.44 mA. Such modeling studies may be useful for personalization of TES devices for finding optimal positioning of electrodes with respect to target and stimulation amplitude range that minimizes side effects.

Magnetic resonance imaging (MRI) is the most powerful instrument for imaging anatomical structures. One of the most essential components of the MRI scanner is a radiofrequency (RF) coil. It induces resonant phenomena and receives the resonated RF signal from the body. Then, the signal is computed and reconstructed for MR images. Therefore, improving image quality by increasing the receiver's (Rx) efficiency is always remarkable. This research introduces a flexible and stretchable receive RF coil embedded in a dielectric-loaded material. Recent studies show that the adaptable coil can improve imaging quality by flexing and stretching to fit well with the sample's surface, reducing the spatial distance between the load and the coil. High permittivity dielectric material positioned between the coil and phantom was known to increase the RF field distribution's efficiency significantly. Recent studies integrating the high dielectric material with the coil show a significant improvement in signal-to-noise ratio (SNR), which can improve the overall efficiency of the coil. Previous research also introduced new elastic dielectric material, which shows improvement in uniformity when incorporated with an RF coil. Combining the adaptable RF coil with the elastic dielectric material has the potential to enhance the coil's performance further. The flexible dielectric material's limitations and unknown interaction with the coil pose a challenge. Thus, each component was integrated into a simple loop coil step-by-step, which allowed for experimentation and evaluation of the performance of each part. The mechanical performance was tested manually. The introduced coil is highly flexible and can stretch up to 20% of its original length in one direction. The electrical performance was evaluated in simulations and experiments on a 9.4T MRI scanner compared to conventional RF coils.

Magnetic Resonance Imaging has become an increasingly reliable source of medical imaging to obtain high quality detailed images of the human anatomy. Application specific coil or an array of coils when placed closely to the anatomy produces high quality image due to the improved spatial signal to noise ratio. Elastic RF coils have been shown to conform to the shape of the patient’s body and drastically reduce the gap between coil and anatomy. First, a major challenge faced by these elastic RF coils is the changing impedance condition as the coil takes a different shape for every individual. Next, an area that could benefit from the improved image quality and patient comfort that comes from flexible RF coil design is endorectal prostate imaging. Demonstrated in the first part of this dissertation is a modular solution to compensate the impedance mismatch. Standalone Wireless Impedance Matching (SWIM) system is an automatic impedance mismatch compensation system that can function independently of the MR scanner. The matching network consists of a capacitor array with RF switches to electronically cycle through different input impedance conditions. The SWIM system can automatically calibrate an RF coil in 3s with a reflection coefficient of less than -15dB resulting in improved Signal-to-noise ratio (SNR) of the sample image by 12% - 24%, based on sample size, when compared to a loaded coil without retuning. For the second part, we propose a novel elastic and inflatable RF coil integrated with the SWIM system for endorectal prostate imaging at 9.4T. A silicone polymer substrate filled with liquid metal alloy is designed and fabricated with a cavity to create ii inflation. This inflatable RF coil is combined with the SWIM system to automatically tune and match after inflating the RF coil for individual levels of inflation. The imaging results have shown a ~10%, ~19%, and ~25 % increase in SNR due to inflation of RF coil at different ROIs in the acquired image. Overall, the methods proposed and discussed in this thesis are a step towards a new generation of RF coil systems for both existing applications and upcoming ones.

It is hypothesized that changes in brain tissue microstructure, particularly degradation of neurites (i.e,. axons and dendrites) and synapses, are early drivers of Alzheimer's disease (AD) pathogenesis. Quantitative magnetic resonance imaging (MRI) tools like diffusion tensor imaging (DTI) have long been used to study AD pathogenesis. Using DTI metrics, structural insights of neuro tissue can be inferred but not directly measured. DTI has proven to be an effective tool indicating fractional anisotrophy (FA) differences between groups of varying AD risk factor, but it does not explicitly quantify pathophysiologically-relevant features like neurite density and complexity. This study aims to develop and validate an advanced diffusion MRI acquisition and biophysical modeling platform that can be used to explicitly quantify changes to brain tissue microstructure, specifically neurite density and complexity. Ultimately, this platform will be used to study the pathogenic mechanisms that drive AD in the pre-clinical and clinical setting.

In the realm of biosensors and nanotechnology, deoxyribonucleic acid (DNA) nanosensors have demonstrated tremendous potential across diverse real-world applications, from environmental monitoring to healthcare diagnostics. Fabrication of nanosensors allows assembling and designing of DNA molecules at nanoscale with high precision and versatility. Such fabricating DNA nanosensors are quite time consuming. Hence it is important to store them in batches. However synthetic DNA molecules can be prone to degradation over time, especially when exposed to various environmental factors like light, heat, or any other chemical contaminants. To address this issue, a shelf life study of DNA nanosensors using various lyoprotectant conditions was carried out to determine the long term stability of such sensors. This study involves fabrication of the dendritic, double - stranded DNA nanosensors involving five strands L1 through L5 conjugated with pHAb fluorophores via N-hydroxysuccinimide ester reaction and acetylcholinesterase (AChE) enzyme, a core component of the sensor. This sensor was originally a fluorescent ACh-selective nanosensors designed to accommodate the BTX ligand, AChE to image the ACh release in the submandibular region of the living mice to report real time quantitative endogenous ACh release triggered by electrical stimulation. AChE enzyme is a good receptor to detect acetylcholine release in the Peripheral Nervous System (PNS). The primary objective of the study was to assess DNA nanosensors with AChE, however due to the intricate interactions, non-specific binding and cost-effectiveness, the shelf life study was carried out separately. The shelf study includes observing DNA nanosensors with different disaccharide lyoprotectants like trehalose and sucrose that were analyzed under different temperature conditions: room temperature (25ºC) and at 50 ºC for different time intervals, over a week time. Also, Observing AChE with various protectants under 50 ºC with and without lyoprotectants for various time intervals like 24 hours and 48 hours. To replicate the real-world transit scenarios, the study also involves test-shipment of the samples with lyoprotectants for 2-3 days to both cross-country and local (in-state). As a result, the use of lyoprotectants, particularly trehalose, has proven to be more resilient and effective in preserving the stability and integrity of DNA nanosensors and Acetylcholinesterase (AChE) enzymes

Over the past few decades, medical imaging is becoming important in medicine for disease diagnosis, prognosis, treatment assessment and health monitoring. As medical imaging has progressed, imaging biomarkers are being rapidly developed for early diagnosis and staging of disease. Detecting and segmenting objects from images are often the first steps in quantitative measurement of these biomarkers. While large objects can often be automatically or semi-automatically delineated, segmenting small objects (blobs) is challenging. The small object of particular interest in this dissertation are glomeruli from kidney magnetic resonance (MR) images. This problem has its unique challenges. First of all, the size of glomeruli is extremely small and very similar with noises from images. Second, there are massive of glomeruli in kidney, e.g. over 1 million glomeruli in human kidney, and the intensity distribution is heterogenous. A third recognized issue is that a large portion of glomeruli are overlapping and touched in images. The goal of this dissertation is to develop computational algorithms to identify and discover glomeruli related imaging biomarkers. The first phase is to develop a U-net joint with Hessian based Difference of Gaussians (UH-DoG) blob detector. Joining effort from deep learning alleviates the over-detection issue from Hessian analysis. Next, as extension of UH-DoG, a small blob detector using Bi-Threshold Constrained Adaptive Scales (BTCAS) is proposed. Deep learning is treated as prior of Difference of Gaussian (DoG) to improve its efficiency. By adopting BTCAS, under-segmentation issue of deep learning is addressed. The second phase is to develop a denoising convexity-consistent Blob Generative Adversarial Network (BlobGAN). BlobGAN could achieve high denoising performance and selectively denoise the image without affecting the blobs. These detectors are validated on datasets of 2D fluorescent images, 3D synthetic images, 3D MR (18 mice, 3 humans) images and proved to be outperforming the competing detectors. In the last phase, a Fréchet Descriptors Distance based Coreset approach (FDD-Coreset) is proposed for accelerating BlobGAN’s training. Experiments have shown that BlobGAN trained on FDD-Coreset not only significantly reduces the training time, but also achieves higher denoising performance and maintains approximate performance of blob identification compared with training on entire dataset.

Glioblastoma brain tumors are among the most lethal human cancers. Treatment efforts typically involve both surgical tumor removal, as well as ongoing therapy. In this work, we propose the use of deuterium magnetic resonance imaging (MRI) to delineate tumor boundaries based on spatial distributions of deuterated leucine, as well as resolve the metabolism of leucine within the tumor. Accurate boundary identification contributes to effectiveness of tumor removal efforts, while amino acid metabolism information may help characterize tumor malignancy and guide ongoing treatment. So, we first examine the fundamental mechanisms of deuterium MRI. We then discuss the use of spin-echo and gradient recall echo sequences for mapping spatial distributions of deuterated leucine, and the use of single-voxel spectroscopy for imaging metabolites within a tumor.

This work focuses on qualifying the performance of an optoelectrical measurement system designed to analyze ribonucleic acid (RNA) within a micro sample. The system is capable of measuring light intensity converted to voltage versus time and is a fast, inexpensive, and portable method for rapid detection of biologics such as SARS-CoV-2 virus, or Covid-19 disease. The measurement system consists of a microfluidic chip and a point of care fluorescent reader.The intent of this research is to measure consistency and robustness of the fluorescent reader combined with the microfluidic chip. The consistency and the robustness of the fluorescent reader within the duty cycle of the system power and the measurement system were analyzed with Six Sigma methods. Control charts, analysis of variance (ANOVAs), and variance components calculations were implemented to characterize the reader system. Through the process of this analysis, baseline characteristics were measured and documented providing valuable data for the improved instrument design. The existing microfluidic chip is a prototype that works in combination with the reader based on fluorescent detection. Baseline studies were required to define any issues related to microfluidic autofluorescence. Multiple designs were tested to measure reduction in autofluorescence in the microfluidics. It was found that certain designs performed better than others. One approach for improvement in the microfluidic chip may be achieved by characterizing and source controlling materials, optimizing layers, mask apertures, and mask orientations to determine reliability in the measurable output through the fluorescent reader. Since the reader and the microfluidic are designed to work together, any future studies should explore testing where the two components are considered a coupled system.

Magnetic resonance imaging (MRI) is a noninvasive imaging modality, which is used for many different applications. The versatility of MRI is in acquiring high resolution anatomical and functional images with no use of ionizing radiation. The contrast in MR images can be engineered by two different mechanisms with imaging parameters (TR, TE, α) and/or contrast agents. The contrast in the former is influenced by the intrinsic properties of the tissue (T1, T2, ρ), while the contrast agents change the relaxation rate of the protons to enhance contrast. Contrast agents have attracted a lot of attention because they can be modified with targeting groups to shed light on some physiological and biological questions, such as the presence of hypoxia in a tissue. Hypoxia, defined as lack of oxygen, has many known ramifications on the outcome of therapy in any condition. Hence its study is very important. The standard gold method to detect hypoxia, immunohistochemical (IHC) staining of pimonidazole, is invasive; however, there are many research groups focused on developing new and mainly noninvasive methods to investigate hypoxia in different tissues.Previously, a novel nitroimidazole-based T1 contrast agent, gadolinium tetraazacyclododecanetetraacetic acid monoamide conjugate of 2-nitroimidazole (GdDO3NI ), has been synthesized and characterized on subcutaneous prostate and lung tumor models. Here, its efficacy and performance on traumatic brain injuries and brain tumors are studied. The pharmacokinetic properties of the contrast agent the perfusion properties of brain tumors are investigated. These results can be used in personalized therapies for more effective results for patients. Gadolinium (Gd), which is a strongly paramagnetic heavy metal, is routinely and widely used as an MR contrast agent by chelation with a biocompatible ligand which is typically cleared through the kidneys. While widely used, there are serious concerns for patients with impaired kidney function, as well as recent studies showed Gd accumulation in the bone and brain. Iron as a physiological ion is also capable of generating contrast in MR images. Here synthesis and characterization of an iron-based hypoxia targeting contrast agent is proposed to eliminate Gd-related complications and provide a cheaper and more economical alternative contrast agent to detect hypoxia.