Spatio-temporal variation in action intervals during ventricular fibrillation (VF) suggestthat the excitable gap may also be distributed spatio-temporally. The observation leadus to hypothesize that distributed pacing can be used to modify and entrain electricalactivation during VF. We tested this hypothesis using simulated VF and animal studies. We simulated VF in a 400 by 400 cell matrix. Simulation results showed that activationpattern could be entrained using spatially distributed stimulation. Up to a certain limit,increasing stimulus strength and density led to improved entrainment. Best entrainmentwas obtained by pacing at a cycle length similar to the intrinsic cycle length. In order to verify whether activation could be entrained experimentally, eight opticallyisolated biphasic TTL addressable stimulators were fabricated. Distributed stimulationwas tested during electrically induced VF in two canines and two swine. Resultsshowed that electrical activation could be entrained in both species. Similar to that insimulation, better entrainment was obtained with denser pacing distribution and atpacing cycle length similar to the intrinsic cycle length. As expected, entrainment wasaffected by tissue thickness. Our results show that spatio-temporally distributed pacingstrength stimuli can be used to modify activation patterns during VF.

In silico models of intracranial aneurysm (IA) evolution aim to reliably represent the mechanical blood flow environment, the biology of the arterial wall and, crucially, the complex link between the two, namely the mechanobiology of healthy and diseased arteries. The ultimate goal is to create diagnostic tools for personalized management and treatment of aneurysm disease. Towards that target, the work presented in this thesis aims to establish a directly interactive link between experimental (in vivo and in vitro) and computational work for biologically and clinically relevant research on aneurysm disease. Mechanobiological hypotheses were firstly investigated in a novel 1D mathematical conceptual model of aneurysm evolution: for the first time these included representations of endothelial heterogeneity and smooth muscle cell (SMC) active stress response and apoptosis. The 1D investigations analysed and assessed the role of wall shear stress (WSS) homeostasis in elastin degradation, and the evolving role of the adventitia as a protective sheath in health and primary load-bearer in disease. The 1D framework was applied to a specific patient's aneurysm using both imaging and histological data to parameterise the model, calculating a material parameter for the adventitital collagen. The predicted evolution captured aspects of tissue changes measured with time focusing on the remodelled tissue wall thickness consistent with the experimental measurements, and physiological cyclic deformation in order to propose an approach to modelling adventitia's adaptive role to load bearing. Furthermore, an existing Fluid-Solid-Growth (FSG) computational framework was adapted and calibrated for the same patient-specific case with the help from the experimental data and the analysis from the 1D framework. This FSG model quantifies the arterial mechanical environment and captures the mechanical response of the fibrous arterial constituents. Comparing 1D and 3D investigations to establish consistency for our models, the 3Dmodel tested the hypothesis of WSS homeostasis, additionally introducing the element of spatial heterogeneity in the definition, and a new hypothesis linking cyclic deformation with collagen growth that ensures a physiological mechanical environment in stabilised aneurysms. Moreover, the FSG framework was applied in a specific rabbit aneurysm case and extended to link growth and remodeling to the detailed representation of the pulsatile blood flow mechanical environment. This research illustrates the power of computational modelling when coupled with rich data sets on the physiology, histology and geometry of healthy and diseased vascular tissue. In particular, the integrative modelling framework provides the foundation for establishing mechanobiological links crucial to aneurysm progression, and a basis for further research towards creating reliable aneurysm clinical tools.

First, an in vitro growth experiment was performed to test the hypothesis that applying dynamic unconfined compression during culture produces het- erogeneous remodeling in newborn bovine articular cartilage explants. Het- erogeneous measures of cartilage microstructure were obtained by biochemical assays and quantified polarized light microscopy. Significant differences were measured between the GAG content in the inner and outer portions of the sam- ples stimulated with dynamic unconfined compression. The COL fiber network was found to be more highly aligned in the inner portion of the sample than in the peripheral region. Next, a poroelastic finite element model with a remodeling subroutine was developed to test the hypothesis that the magnitude of relative interstitial fluid velocity and maximum principle strain stimulate GAG and COL fiber network remodeling, respectively, in articular cartilage during culture with dynamic unconfined compression. The GAG remodeling law was successful in predicting the heterogeneous changes in GAG content. The collagen remodeling law was not successful in predicting the changes in the COL network microstructural orientation, suggesting another mechanical cue is responsible for stimulating the remodeling of the COL fiber network.

<p>Breast cancer is one of the most commonly diagnosed cancers in women, with 1 in 8 women diagnosed during her lifetime. Distant metastatic breast cancer accounts for a majority of deaths in breast cancer patients. Changes in both the architecture and the biochemical composition of the extracellular matrix (ECM) occur during metastatic dissemination at both the primary tumor and the early metastatic niche. These changes play a significant role in the cell fate, and can alter proliferation, migration, and quiescence of cancer cells. This study utilizes tandem mass spectrometry to study ECM protein changes, specifically in the lungs, using an immune competent murine model of metastatic breast cancer. Liquid chromatography-tandem mass spectrometry was used to identify and quantify key ECM proteins in the primary tumors, lungs, and metastatic tumors during cancer progression. Fibronectin (FN) was upregulated in the primary tumor, suggestive of a more invasive mesenchymal-like cell. However, FN was decreased in abundance in metastatic tumors, which is favorable for a more epithelial phenotype, prompting tumor growth. The diseased lungs appear to have highly collagenous proteins, suggesting an increased stiffness in the matrix. This increase in stiffness would reduce physiologically induced strains, and potentially facilitate growth of metastatic lesions in the lungs. Characterization of the changes in the ECM during cancer progression will aid in development of future therapies as well as guide the design of relevant <i>in vitro</i> models, ultimately enhancing the knowledge of this phenomenon.</p>

<p>The recent FDA regulatory clearance for the 7 tesla Magnetic Resonance Imaging (MRI) system has led to increased interest in clinical ultra-high field (UHF) applications. However, to robustly achieve the expected increase in signal-to-noise ratio (SNR) at UHF, the radiofrequency (RF) challenges need to be met, namely, problems with higher RF power, worse <i>B₁⁺</i> inhomogeneity (signal voids) and increased tissue dielectric properties at higher frequency, all of which usually results in increased specific absorption rate (SAR). The parallel transmission (pTx) techniques are generally accepted as a realistic solution, providing improvement in the <i>B₁⁺</i> homogeneity with good RF efficiency while reducing peak local SAR. We designed a hybrid circuit-spatial domain optimization to accelerate the design of a double row pTx head coil. The method predicted consistent coil scattering parameters, component values and <i>B₁⁺</i> field. RF shimming of the calculated field maps matched in vivo performance. To further increase the <i>B₁⁺</i> homogeneity in tissue, we added high dielectric material (HPM) pads near the coil, as the displacement currents in the HPM induced secondary <i>B₁⁺</i> in tissue. This raises a RF safety question of how to monitor millions of local SAR (complex valued Q-matrix) in the tissue voxels, for any weightings (forward voltages) applied to the pTx system. We implemented VOPs based on singular value decomposition to compress the Q-matrices with a compression ratio >100, effectively monitoring the maximum peak local SAR values at given weighting amplitudes.</p>

The standard electrical stimulation waveform used for electrical activation of nerve is a rectangular pulse or a charge balanced rectangular pulse, where the pulse width is typically in the range of ∼100 µsec through ∼1000 µsec. In this work, we explore the effects of a continuous sinusoidal waveform with a frequency ranging from 5 through 20 Hz, which was named the Low Frequency Alternating Current (LFAC) waveform. The LFAC waveform was explored in the Bioelectronics Laboratory as a novel means to evoke nerve block. However, in an attempt to evoke complete nerve block on a somatic motor nerve, increasing the amplitude of the LFAC waveform unexpectedly produced nerve activation, and elicited a strong non-fatiguing muscle contraction in the anesthetized rabbit model (unpublished observation). The present thesis aimed to further explore the phenomenon to measure the effect of LFAC waveform frequency and amplitude on nerve activation.<div><br></div><div>In freshly excised canine cervical vagus nerve (n=3), it was found that the LFAC waveform at 5, 10, and 20 Hz produced burst modulated activity. Compound action potentials (CAP) synchronous to the stimuli was absent from the electroneurogram (ENG) recordings. When applied <i>in-vivo</i>, LFAC was capable of activating the cervical vagus nerve fibers in anaesthetized swine (n=5) and induced the Hering-Breuer reflex. Additionally, when applied <i>in-vivo</i> to anesthetized Sprague Dawley rats (n=4), the LFAC waveform was able to activate the left sciatic nerve fibers and induced muscle contractions.</div><div><br></div><div>The results demonstrate that LFAC activation was stochastic, and asynchronous to the stimuli unlike conventional pulse stimulation where nerve and muscle response simultaneously and synchronously to stimulus. The activation thresholds were found to be frequency dependent. As the waveform frequency increases the required current amplitude decreases. These experiments also implied that the LFAC phenomenon was most likely to be fiber type-size dependent but that more sophisticated exploration should be addressed before reaching clinical applications. In all settings, the LFAC amplitude was within the water window preventing irreversible electrochemical reactions and damages to the cuff electrodes or nerve tissues. This thesis also reconfirms the preliminary LFAC activation discovery and explores multiple methods to evaluate the experimental observations, which suggest the feasibility of the LFAC waveform at 5, 10, and 20 Hz to activate autonomic and somatic nerve fibers. LFAC appears to be a promising new technique to activate peripheral nerve fibers.</div>

The development of wearable devices to monitor biosignals and collect real-time data from biological systems at all scales from cellular to organ level has played a significant role in the field of medical engineering. The current coronavirus disease 2019 (COVID-19) pandemic has further increased the demand for remote monitoring and smart healthcare where patient data can be also be accessed from a remote distance. Recent efforts to integrate wearable devices with artificial intelligence and machine learning have transformed conventional healthcare into smart healthcare, which requires reliable and robust recording data. The biomedical devices that are mechanically stretchable and compliant have provided the capability to form a seamless interface with the curvilinear, soft surface of tissues and body, enabling accurate, continuous acquisition of physical and electrophysiological signals. This dissertation presents a comprehensive set of functional materials, design principles, and fabrication strategies to develop mechanically stretchable and compliant biomedical devices tailored for various applications, including (1) a stretchable sensor patch enabling the continuous monitoring of swallowing function from the submental/facial area for the telerehabilitation of patients with dysphagia, (2) a human hand-like sensory glove for advanced control of prosthetic hands, (3) a mechanically compliant manipulator for the non-invasive handling of delicate biomaterials and bioelectronics, and (4) a stretchable sensors embedded inside a tissue scaffold enabling the continuous monitoring of cellular electrophysiological behavior with high spatiotemporal resolution.<br>

The research presented in this thesis focuses on understanding the biomechanical effects of various cranial features that are often ignored in finite element models (FEMs) because their size, position, and complex shapes make them difficult to model. Specifically, this work examines the effects of the alveoli (tooth sockets), periodontal ligament, and squamosal suture on the stress and strain distributions in a cranium under masticatory and dynamic tooth loads. Results from this research will help determine if these features have a significant effect on stress and strain patterns and will yield guidelines as to if or under what conditions they need to be modeled in future FE skull model analyses. As part of this research, three sets of FEMs were developed to address a hypothesis focusing on each cranial feature. The first set of models examined the effect of the tooth sockets on the stress and strain distributions in a cranium under static biting conditions to determine if improperly modeled sockets produce strong global effects in craniofacial regions. The second set of models were used to assess the effect of the PDL's material behavior on the stresses and strains in a cranium under static biting and dynamic tooth loading conditions to determine if the PDL plays an important role in reducing stresses and strains in a model. The final set of models were used to determine the effect of the squamosal suture size on the stresses and strain energies in a cranium under static biting conditions to see if an increase in suture size decreases the risk of separation of the temporal bone from the parietal bone. Results for all analyses indicate the effects of the cranial features are local (i.e. within the vicinity of the feature), with no meaningful global effects. This suggests the sockets, PDL, and squamosal suture do not play an important role in global stress and strain distributions in a cranium under masticatory and dynamic tooth loads. Therefore, it may be safe to ignore the sockets, PDLs, and squamosal sutures during the FE modeling process if the objective of the analysis is to understand global stress and strain patterns.

Hydrocephalus is developed when the flow of cerebrospinal fluid is obstructed in the ventricles and a pressure build-up is generated within the brain. Ventricular shunts are used to remove excess fluid from the brain, but these shunts have a common problem of failure due to the shunt being obstructed by the build-up of astrocytes. To address this, two sets of 27 designs of ventricular shunts were identified and analyzed with parameters that could potentially reduce obstruction risks. The performance of these designs was examined using fluid simulations on these two sets of 27 designs. One set explored close-tipped shunt designs, and the other assessed open-tipped ones. Following these simulations, adjustments were made to three design variables of the ventricular catheters - inlet hole size, inner shunt diameter, and inner-segment distance. The goal was to optimize these variables to prevent obstruction, ensuring three key design objectives were met: maintaining wall shear stress, ensuring a balanced inlet and outlet pressure difference, and achieving a uniform flow distribution.

An estimated 27 million Americans suffer from osteoarthritis (OA). Symptomatic OA is often treated with total knee replacement, a procedure which is expected to increase in number by 673% from 2005 to 2030, and costs to perform total knee replacement surgeries exceeded $11 billion in 2005. Subject-specific modeling and finite element (FE) predictions are state-of-the-art computational methods for anatomically accurate predictions of joint tissue loads in surgical-planning and rehabilitation. Knee joint FE models have been used to predict in-vivo joint kinematics, loads, stresses and strains, and joint contact area and pressure. Abnormal cartilage contact pressure is considered a risk factor for incidence and progression of OA. For this study, three subject-specific tibiofemoral knee FE models containing accurate geometry were developed from magnetic resonance images (MRIs). Linear (LIN), Neo-Hookean (NH), and poroelastic (PE) cartilage material models were implemented in each FE model for each subject under three loading cases to compare cartilage contact pressure predictions at each load case. An additional objective was to compare FE predictions of cartilage contact pressure for LIN, NH, and PE material models with experimental measurements of cartilage contact pressure. Because past studies on FE predictions of cartilage contact pressure using different material models and material property values have found differences in cartilage contact pressure, it was hypothesized that different FE predictions of cartilage contact pressure using LIN, NH, and PE material models for three subjects at three different loading cases would find statistically significant differences in cartilage contact pressure between the material models. It was further hypothesized that FE predictions of cartilage contact pressure for the PE cartilage material model would be statistically similar to experimental data, while the LIN and NH cartilage material models would be significantly different for all three loading cases. This study found FE and experimental measurements of cartilage contact pressure only showed significant statistical differences for LIN, NH, and PE predictions in the medial compartment at 1000N applied at 30 degrees, and for the PE prediction in the medial compartment at 500N applied at 0 degrees. FE predictions of cartilage contact pressure using the PE cartilage material model were considered less similar to experimental data than the LIN and NH cartilage material models. This is the first study to use LIN, NH, and PE material models to examine knee cartilage contact pressure predictions using FE methods for multiple subjects and multiple load cases. The results demonstrated that future subject specific knee joint FE studies would be advised to select LIN and NH cartilage material models for the purpose of making FE predictions of cartilage contact pressure.