Organizational Unit:
Wallace H. Coulter Department of Biomedical Engineering

Research Organization Registry ID
Previous Names
Parent Organization
Parent Organization
Organizational Unit
Includes Organization(s)

Publication Search Results

Now showing 1 - 10 of 536
  • Item
    Development and Characterization of Laser-Activated Perfluorocarbon Nanodroplets as Theranostic Agents
    (Georgia Institute of Technology, 2023-11-28) Zhao, Andrew Xin
    Perfluorocarbon nanodroplets have been primarily explored as a next-generation contrast agent for ultrasound imaging, providing significant advantages over microbubbles in terms of circulation time and size, which enables them to extravasate and reach extravascular targets. Furthermore, these nanoparticles can be triggered acoustically or optically to form microbubbles in situ. While acoustic droplet vaporization has been extensively studied, optical droplet vaporization holds promise as an alternative approach. It allows for the incorporation of photoacoustic imaging to visualize vaporization and can be used to vaporize high boiling point cores safely and repeatedly. In this dissertation, optically triggerable PFCnDs are further developed as theranostic agents. The formulation of PFCnDs is varied to gain a better understanding of optical droplet vaporization and to improve the particle’s imaging performance. The therapeutic applications are explored from two angles: impact of the vaporization and release of encapsulated drugs. In CHAPTER 3, the performance of optically activatable PFCnDs are examined, considering factors such as shell and core materials, laser parameters, and environmental conditions. Novel longitudinal behaviors of the nanodroplets are described that can allow for more effective detection in the body and provide insights into how the formulation impacts the vaporization threshold. These findings provide a better foundation for advancing our understanding of optical droplet vaporization. In CHAPTER 4, the effect of lasing the nanodroplets in close proximity to the cells are explored. The nanodroplets exhibit a similar behavior to sonoporation, creating pores in the cell membrane that enable extracellular molecules to enter the cytoplasm. The different conditions required to optimize this behavior and viability are explored in detail such as the concentration of the PFCnDs, laser fluence, pulse number, and size of the molecules that can enter. In CHAPTER 5, double emulsion perfluorohexane nanodroplets (dePFHnDs) are developed by modifying the formulation of PFCnDs to encapsulate hydrophilic molecules. This new class of PFCnDs can be tracked photoacoustically and triggered to release drugs acoustically. Photoacoustic monitoring can be performed without releasing the encapsulated drug, allowing for estimations of release based upon signal reduction. In summary, this dissertation represents an advancement of PFCnDs as a theranostic agent. Our understanding of optical droplet vaporization is improved, allowing for better diagnostic applications, and expand the therapeutic capabilities of these nanodroplets in drug delivery. These developments contribute to making PFCnDs a more effective drug delivery agent.
  • Item
    Modulation of Sphingosine-1-Phosphate Receptor Signaling as a Regenerative Immunotherapy for Volumetric Muscle Loss
    (Georgia Institute of Technology, 2023-11-13) Hymel, Lauren Alexandra
    Extremity trauma is a significant clinical challenge among both civilian and military populations, particularly in cases that result in volumetric muscle loss (VML). Many approaches aimed to treat VML fail to pay attention to the local endogenous immune response of the host which underlies the chronic inflammation and fibrotic signaling characteristic of VML pathology and results in loss of function. Sphingosine-1-phosphate (S1P) is a bioactive sphingolipid locally produced during inflammation through the activity of endogenous sphingosine kinases and is increasingly recognized as a crucial regulator of the immune response. S1P signals through its 5 G protein-coupled receptors (S1PR1-5) to elicit a range of cellular responses governing complex functions such as immune cell trafficking, vascular integrity, and fibrosis. Using a pre-clinical model of VML, we found that critically sized injuries are characterized by a marked increase in the local production of S1P in injured tissue, a sustained presence of M2 macrophages with aberrant TNFa and TGFb co-expression, and significantly increased pro-fibrotic FAPs which highly express b1-integrin and are predisposed towards fibrotic differentiation. We demonstrated that by mediating S1P synthesis, S1P spatial gradients are perturbed which allows for regulation of immune cell trafficking into the injured tissue. We next delivered an S1P chaperone-based protein therapeutic to activate S1PR1 signaling on the damaged endothelium to increase its barrier function following injury. Our findings showed that sustained S1P/S1PR1 vascular signaling reduced pro-fibrotic FAP accumulation in injured muscle and abrogated fibrotic deposition. Finally, we revealed that small molecule antagonism of S1PR3 leads to enhanced skeletal muscle regeneration. To achieve a localized strategy for S1PR3 antagonism, we delivered the pharmacological inhibitor from nanofibrous poly(ethylene glycol)/hyaluronic acid hydrogels which facilitated macrophage egress from the local milieu and improved muscle repair. Taken together, these findings represent an improved understanding of the role of S1P/S1PR signaling in the pathogenesis of VML. Additionally, this work establishes that modulation of S1PR signaling provides an effective therapeutic strategy for shifting the local microenvironment from pro-fibrotic to pro-regenerative following traumatic injury.
  • Item
    Spatio-temporal controlled delivery strategies for the prevention of muscle degeneration after severe rotator cuff tear
    (Georgia Institute of Technology, 2023-09-18) Anderson, Leah E.
    Rotator cuff tear is a significant musculoskeletal injury. Following rotator cuff tendon injury, the corresponding muscle undergoes degenerative changes due to unloading and disuse. The standard of care after severe rotator cuff tear is tendon reattachment, but even after reattachment and associated rehabilitation protocols, the muscle degeneration that occurs after injury is not fully reversed. Furthermore, the degree of muscle degeneration before repair has been shown to correlate to poor surgical outcomes. Thus, there is a need for treatment strategies to reduce or prevent muscle degeneration after severe rotator cuff tear. The long-term goal of this research was to employ treatment and biomaterial delivery strategies that influence the muscle cellular milieu in a spatio-temporal manner, to prevent muscle degeneration after severe rotator cuff tear or repair. In Aim 1, using an acute rotator cuff injury model (no tendon repair), systemic delivery of a bone marrow mobilizing agent, VPC01091, was used to “push” cells into circulation, while the chemokine SDF-1alpha, was locally delivered to injured rat supraspinatus muscle via hydrolytically degradable N-desulfated heparin-based microparticles to “pull” cells to the injury site and increase cell recruitment over previously observed cellular changes due to SDF-1alpha treatment alone. In Aim 2, hydrolytically degradable N-desulfated heparin-based hydrogel fragments were utilized for the spatialized delivery of SDF-1alpha, and thus cellular recruitment, to two distinct regions of rat supraspinatus muscle after rotator cuff tear. Finally, in Aim 3, supraspinatus muscle healing response after tendon repair was explored with and without micronized dehydrated human amnion/chorion membrane (dHACM) treatment in a clinically relevant model of delayed tendon reattachment in rat.
  • Item
    TToward Systems Miniature Optical Microscopy: Instrumentation and Biomedical Applications
    (Georgia Institute of Technology, 2023-08-21) Son, Jeonghwan
    Optical microscopy has offered versatile formats of imaging techniques, accompanied by the rapid development of miniature and open-source approaches, revolutionizing various biomedical applications. However, the adoption and utilization of recent optical imaging systems still face challenges due to the limited accessibility and applicability, restricting the potential applications of advanced optical microscopy. Thus, this thesis focuses on providing instrumentation and applications of recent fluorescence microscopy with high portability, accessibility, and adaptable solution via miniature and open-source approaches. The specific aims involve (1) advancing miniaturized microscopy (miniscopy) for high-resolution imaging of in vivo tissues and neural activity visualization in freely behaving mice, providing real-time insights into tissue dynamics and neural patterns; (2) developing systems for simultaneous in situ cell imaging using array miniscopy on multi-well plates, facilitating the study of cellular behavior and interactions within their native environment; and (3) advancing high-throughput imaging flow cytometry through the implementation of portable light-sheet microscopy, enabling three-dimensional characterization of cellular populations. The thesis strives to foster the development and application of innovative optical microscopy with the ultimate goal of enhancing biomedical research capabilities and facilitating discoveries in the field.
  • Item
    Enabling Accurate Cardiopulmonary Monitoring Using Machine Learning and a Chest-Worn Wearable Patch
    (Georgia Institute of Technology, 2023-07-30) Chan, Michael
    The development of non-invasive instruments to monitor health parameters such as HR, RR, SpO2, VO2, blood pressure, body temperature, etc. have had a significant impact in the history of medicine. Without much pain and infection risk, doctors may gather information about a patient’s body and make better informed make informed decisions to treat the patient. The emergence of these instruments ever since the 19th and 20th century have already saved many lives and revealed many mysteries of our body. Advancements in miniaturization and digitization technologies in recent decades have further encouraged public’s interests to measure health parameters anywhere and anytime, most notably in the form factor of a wrist-worn watch. However, this convenient form factor also presents an inevitable constraint—it cannot capture cardiac vibrations continuously and hence would result in a missed opportunity to examine cardiac function more comprehensively. In contrast, another convenient form factor of a chest-worn wearable patch is a more fitting option. The chest-worn wearable patch has been designed to capture several physiological signals such as ECG, SCG, and PPG simultaneously. However, since this measurement site is not well-perfused and less studied, it remains uncertain whether it is accurate enough for estimating health parameters. Accordingly, in this dissertation, we have developed a series of algorithms to improve and validate the accuracy of the health parameters estimated. To estimate these health parameters, DSP pipelines that consider the sensing principles and the physiology have been employed for denoising, demodulation, and feature extraction while ML models were used capture the complex relationship between the extracted features and the target variables. Conventionally, physiological knowledge was used with DSP pipelines to obtain signal representation that could be easily used with ML models sequentially. More recently, advanced ML models such as those based on DL architectures have emerged and outperformed the conventional methods owning to their expressivity to learn more salient, scalable representations of the signals, together with regression/classification simultaneously and synergistically. Nevertheless, we have noticed that these DL methods were also less interpretable and analytical and hence may not deepen our understanding of the physiological signals, the physiology, and the diseases while they become more powerful and advanced. Note that the term interpretability is referred to the ability to interpret a ML model using engineering concepts rather than to interpret it using clinical terms. To avoid this “hidden trap” of ML, we have reconsidered and tested different ways to combine DSP, physiological knowledge, and ML. Such observation has motivated the three aims presented in this dissertation, each has followed an implicit approach to combine DSP, physiological knowledge, and ML for the estimation tasks. In the first aim, we have constructed the classical DSP pipelines for deriving health parameters and replaced one functional module with ML models for SpO2 estimation and RR estimation. In the second aim, we have replaced multiple functional modules with ML models for RR estimation and HR estimation. In the last aim, we have designed a DL architecture with DSP-inspired functional modules. Collectively, we have developed algorithms that harness the interpretability of DSP, leverage the flexibility, expressive power of DL, and exploit the data available at hand and the physiological knowledge through MTL. Overall, this work has addressed the accuracy problem of the chest-worn wearable patch by proposing a new algorithmic direction for merging DSP, physiological knowledge, and ML in a cohesive manner.
  • Item
    Experimental and Computational Analysis of Pathogen Emergence and Antibiotic Resistance in a Cystic Fibrosis Airway Infection Model
    (Georgia Institute of Technology, 2023-07-24) Davis, Jacob Dylan
    The human body harbors at least twice as many bacteria as it does human cells. Most of these bacteria are harmless, but the emergence of pathogens is common in many human body systems. Treatment of these infections is often with antibiotics, which can have non-target effects and remove protective flora from the body. This project was designed to create a model system of airway bacterial communities that is amenable to the development of effective experimental and computational investigations that shed light on pathogen emergence and antibiotic resistance. For the experimental analysis, we transformed three bacterial species found in human airways with the goal of making them easily quantifiable with available microscopic and spectrophotometric techniques. The bacteria were grown dynamics were studied, as well as the effects of pH on the system. Community resistance to a beta-lactam was studied by tracking the hydrolyzation of the antibiotic by non-targeted species, showing that non-focal species are important to consider when choosing an antibiotic treatment. To quantify interactions among the different species, mathematical models within the Lotka-Volterra framework were developed and parameterized. The existing framework was then expanded to incorporate antibiotic and metabolic data in the community model. Although the community size of the model system is small - to allow for comprehensive data generation - this experimental and mathematical system constitutes a prototype for investigating larger models that can be used to predict how pathogens survive in different communities and under altered environmental conditions and antibiotic treatments.
  • Item
    Measurement and Prevention of Occlusive Arterial Thrombosis
    (Georgia Institute of Technology, 2023-05-19) Bresette, Christopher
    Arterial thrombosis is the process of forming a blood clot in an artery and is a leading cause of ischemic events such as heart attacks and strokes. These ischemic events account for 20% of all deaths in the United States. In order to reduce the mortality from ischemic events we need better methods of predicting when arterial thrombi will form, new drugs that can prevent arterial thrombosis without increasing the risk of bleeding, and basic research into how large occlusive clots are structured and formed. There is a large need for a device that can quantify a patient’s risk of arterial thrombosis based on a blood samples. However, previously developed assays fail to include the critical variables required for arterial thrombosis and are unable to reliably aid clinicians in predicting future events or making patient-specific changes in treatment for secondary prevention. A point-of-care (POC) test, called Thrombocheck, that recreates arterial thrombosis was built by incorporating the 3 key features of arterial thrombosis: high shear, a thrombogenic surface, and platelets and vWF. In addition to incorporating the 3 key features of thrombosis, the Thrombocheck has the advantage that it does not require chemical agonists to induce clotting and forms large, stable thrombi. We show that the Thrombocheck is a functional test for high shear clot formation, is sensitive to anti-platelet use and provides a unique endpoint for arterial thrombosis. Current anti-thrombotic medications used to treat patients at high risk of forming arterial thrombi reduce the likelihood of forming arterial thrombi but also cause an increase in bleeding-related mortality. While pharmaceutical therapies for arterial thrombosis have historically acted through preventing platelet activation, it has been hypothesized that targeting the shear-sensitive von Willebrand Factor (vWF) could allow for prevention of arterial thrombosis without adverse effects on coagulation and low-shear hemostasis. N-acetylcysteine (NAC) is an existing drug currently used for treating acetaminophen overdose and cystic fibrosis which also cleaves vWF. By breaking the disulfide bonds that form the backbone of vWF, long vWF multimers are chopped into smaller, less active multimers. It has been used to shorten vWF, reduce platelet binding in a population with abnormally long vWF and lyse platelet aggregates. We show that in a healthy population N-acetylcysteine (NAC) prevents arterial thrombosis in a dose dependent manner, increasing occlusion times at concentrations of 3–5 mM and completely preventing platelet aggregation at concentrations at or above 10 mM. An in vivo murine model shows that the effect of NAC on arterial thrombosis is lasting, cumulative and does not increase bleeding time. NAC can therefore be repurposed as an anti-thrombotic, preventing arterial thrombosis without affecting bleeding. Additional research into the basic science of platelet aggregation directs future studies attempting to predict and prevent thrombosis. While many studies have focused on single platelet attachments or small microfluidic aggregates, it is critical to understand how the aggregation of billions of individual platelets leads to the creation of a large, occlusive clot. These clots have a unique structure compared to coagulation clots at both the small aggregate level (1-10 um) and at the whole clot level (100-1,000 um). At the aggregate level arterial thrombi are extremely porous, while at the whole clot level they have increased mechanical strength and have striations which indicate the presence of a large-scale architecture (~100 um). In this work, large-scale clot architecture was characterized in multiple experimental systems and shown to appear as periodic clot “fingers" that grow perpendicular to the direction of flow. At occlusion, these distinct structures merge to form a single clot. A parsimonious model comprised of 4 platelet behaviors was able to recreate this large-scale architecture and to predict how clot structure changes based on initial conditions. By defining large-scale clot architecture and modeling its formation, we have a better understanding of the processes leading to thrombosis formation.
  • Item
    Metabolic and Bioelectric Crosstalk in Directed Differentiation and Spatial Patterning of iPSC-Derived Cardiomyocytes
    (Georgia Institute of Technology, 2023-05-08) Norfleet, Dennis Andre
    The goal of multi-cellular engineered living systems is the design and manufacturing of multicellular systems with novel form or function using engineering design principles. Induced pluripotent stem cells represent an excellent tool to enable actualization of these design goals because of their intrinsic pluripotent capacity and recapitulation of various embryogenesis and organogenesis processes. The objective of this research was to investigate through computational modeling how molecular components of bioelectric and metabolic systems alter multicellular bioelectric patterning and cell metabolic flux dynamics, and to extend system understanding to guide emergent morphogenic outcomes via external modulation of the culturing environment. The central hypothesis of this work was that specific media compositions can alter molecular components of bioelectric and metabolic multicellular systems in a predictable manner, leading to desired morphologies, cell phenotypes, and novel functionalities. In the first study, a multiscale bioelectric computational model describing human iPSC tissue-scale membrane voltage potentials (Vmem) was developed to understand unexplored patterning outcomes when various molecular components of the bioelectric system are altered by culture media. Model simulations accurately predicted multicellular Vmem patterns when one or more molecular components were altered, as quantitatively confirmed by a machine learning-based quantitative image pattern similarity analysis. In the second modeling analysis, a genome-scale computational model of the human metabolic network was expanded with additional descriptors to investigate how induced pluripotent stem cells reroute metabolic fluxes and achieved cell growth objectives during cardiomyocyte differentiation under various culture media compositions. This framework integrated transcriptomic, thermodynamic, kinetic, and proteomic and novel fluxome constraints including transport exchange between the cytosol and extracellular environment. From a comparative analysis across multiple published studies and our own experimental validations, we observed that the combination of novel and previous model constraints was required to replicate experimental media-induced changes in metabolic network dynamics during pluripotency and hiPSC-cardiomyocyte (hiPSC-CM) differentiation. We extended this study to a novel media supplementation condition of glutamine and ascorbic acid and found that experimental extracellular flux assays supported the model-predicted improvements to metabolic respiration of iPSC-derived cardiomyocyte progenitor cells. In summary, these results collectively validate the potential for model-guided media design of engineered living systems using understanding of bioelectric and metabolic systems properties.
  • Item
    Novel Explainability Approaches for Analyzing Functional Neuroinformatics Data with Supervised and Unsupervised Machine Learning
    (Georgia Institute of Technology, 2023-05-01) Ellis, Charles Anthony
    The use of artificial intelligence in healthcare is growing increasingly common. However, the use of artificial intelligence within the context of neuropsychiatric settings with functional neuroinformatics modalities like electroencephalography (EEG) and functional magnetic resonance imaging (fMRI) is still in its infancy. While there are a number of obstacles preventing the development of diagnostic and prognostic tools for clinical neuroinformatics, a key obstacle is the lack of explainability approaches uniquely adapted to the field. This lack of explainability methods has implications both within a clinical context and within a biomedical research context. In this dissertation, we propose a series of novel explainability approaches for systematically evaluating what deep learning models trained for both classification and clustering have learned from raw EEG data. These explainability approaches provide insight into key spatial, spectral, temporal, and interaction features uncovered by models. Within the context of fMRI, this dissertation expands upon existing explainability approaches for neuroimaging classification by combining them with approaches that estimate the degree of model confidence in predictions. Additionally, this dissertation presents several novel analyses that can be applied to supervised fMRI classifier explanations to gain insight into models. This dissertation further presents several novel explainability approaches for insight into both hard and soft clustering algorithms applied to fMRI data.
  • Item
    General and interpretable models for inferring dynamical computation in biological neural networks
    (Georgia Institute of Technology, 2023-04-30) Sedler, Andrew R.
    The sequential autoencoder (SAE) has proven to be a powerful deep architecture for modeling large-scale electrophysiological recordings, with a variety of applications in neuroscience and neural engineering. In particular, latent factor analysis via dynamical systems (LFADS) explicitly models latent dynamics and inputs to infer firing rates with state-of-the-art accuracy. However, effectively fitting such models requires time consuming and resource-intensive hyperparameter tuning with reference to supervisory information (e.g. behavioral data) for each new dataset. Additionally, the conditions under which the dynamics learned by the SAE faithfully reflect those of the underlying biological system are unclear and under-explored, limiting the insights that can be gained from internal components of the model. The objectives of this thesis are to simplify training of deep, dynamics-based neural population models on binned spiking activity and to improve the interpretability of the latent dynamics they learn. The first aim of this research was to develop a framework for robust training of SAEs on neural data. In Chapter 2, we present a novel regularization strategy and an efficient hyperparameter tuning approach which allowed us to reliably obtain high-performing models on a wide variety of datasets. The second aim was to evaluate and improve the interpretability of the latent dynamics learned by SAEs. In Chapter 3, we show that widely-used recurrent neural networks (RNNs) struggle to accurately recover dynamics from synthetic datasets, and that neural ordinary differential equations solve many of these issues. The last aim was to address several of the practical challenges of training and applying SAEs in neuroscience. Finally, we describe two open-source projects: a simple, modular, and extensible implementation of LFADS (Chapter 4) and a deployment framework that makes it easier to leverage managed infrastructure for large-scale training (Chapter 5).