Development of a Reliable Drive System for Medical Ultrasound Imaging
Shipra Trivedi, Aarthee Baskaran, Sarah Abourakty, Sarah Ames
Advised by: Zachary Leonard (Rivanna Medical Inc.)
Rivanna Medical, Inc. is dedicated to improving global healthcare through the development of world-first, imaging-based medical technologies including Accuro, which provides automated, real-time ultrasound image guidance for epidural injections. The goal of this capstone project was to develop a reliable electromechanical drive system for a new medical ultrasound imaging device. The project scope included the development and implementation of test protocols to evaluate the reliability of the drive system. Additionally, the team designed and assembled test fixtures and a proof-of-concept design alternative for a flex circuit management system. Rivanna Medical will use the results of this project to guide the design and improve the reliability of new products.
ML-based derivation of blood input function for human brain FDG-dPET
Gabe Hyman, Colton Bogucki
Advised by: Bijoy Kundu (Radiology and Medical Imaging)
To facilitate quantitative analysis of dynamic fluoro-2-deoxy-D-glucose positron emission tomography (dFDG-PET) in human brain, researchers are currently required to perform time-consuming and imprecise manual annotations on the scan in order to localize the carotid arteries. This work aims to demonstrate the ability to automate this annotation process and successfully develop a Machine Learning (ML) based model to correctly recognize and segment out these regions of interest automatically.
Computational Flow Dynamics Analysis of Pelvic and Abdominal Veins Using CT, Venography, and Duplex Imaging
Jessica Cornthwaite, Katherine Byrd, Katherine Dunn
Advised by: John Angle (Interventional Radiology), Haibo Dong (Mechanical and Aerospace Engineering)
The goal of this project is to analyze data from CT, duplex ultrasound, venography, intravascular ultrasound, and pressure measurements from patients with DVT to develop a CFD model of the lower extremity venous anatomy and to understand the contribution of flow rates and patterns on the formation of DVT. In order to accomplish this goal, we will be reconstructing CT data of the iliac vein with volume rendering to create a 3D model. This model will then be used to run simulations of blood flow. To run these simulations, we must identify the appropriate and most significant parameters to consider. We will then verify our model with clinical data as well as analyze blood flow in patients with DVT and healthy patients. Modeling individual patient anatomy of patients will potentially allow us to determine why stent stenosis occurs more frequently in the lesser curve of the iliac vein.
Evaluation of Caffeic Acid Transport in an In Vitro 3D Model of Hepatocellular Carcinoma
Caroline Doyle, Ella Frazier, Lillian Way
Advised by: Luke Wilkins (Vascular & Interventional Radiology), David Brautigan (Microbiology, Immunology & Cancer Biology)
Hepatocellular carcinoma (HCC) is the fifth most common tumor world-wide, and cases continue to rise creating a pressing need for a more effective treatment. Previous research by our advisors has shown that the phytochemical, caffeic acid (CA), when paired with a traditional transarterial embolization (TAE) procedure, causes more extensive reduction in HCC tumor volume when compared to a TAE procedure alone. However, the transport and distribution of CA within an HCC tumor is currently difficult to determine in vivo. In this project, we designed an in vitro 3D Transwell model of HCC to assay for transport of solutes that include a tracking dye, labeled protein, and different formulations of CA. The results will serve as a tool for the optimization of improved TAE procedures for future animal and human clinical trials.
Assessment of Hydrogels to Aid in Point of Care Ultrasound Barrier Reduction at the University of Virginia
Arghya Shetty, Deborah Wood
Advised by: Masahiro Morikawa (Family Medicine)
Point of care ultrasound (POCUS) describes the practice of trained physicians obtaining and interpreting ultrasound images on a patient to diagnose medical problems at the bedside. Though emergency medicine physicians widely use POCUS, primary care physicians can use this practice as an alternative to the stethoscope to provide more timely diagnoses. However, many institutional barriers exist for POCUS at UVA, such as the lack of machines, funding, and a governing body to oversee ultrasound image storage. Our project focused on how physicians can efficiently implement POCUS to incentivize addressing these barriers.
Since ultrasound gel is messy and single-use, we assessed how well reusable accessible hydrogels behave as replacement acoustic couplants.
Developing Preliminary Point-of-Care-Ultrasound Competency Guidelines for Internal Medicine
Andrew Porter, Emma Stiglitz
Advised by: Masahiro Morikawa (Family Medicine)
Point-of-care-ultrasound (POCUS) technology, including handheld models, are portable ultrasound devices that are cost-effective and similarly capable to traditional models. However, they have failed to be adopted into regular clinical practice in Internal Medicine due in part to a lack of standard training guidelines from the professional societies. This project aims to design preliminary guidelines, in the form of a handbook, that includes set modules and checkpoints for physicians to follow. This collection of essential skills can be used to develop a more robust and standardized certification curriculum. Thus, eliminating the gray area that currently exists around the current qualification standards necessary for using POCUS in practice.