Solid tumors like liver cancer will grow by promoting angiogenesis, the development of new blood vessels. These vessels are disordered compared to normal vasculature, leading to spatial-temporal differences in blood perfusion to the tumor. Cancer therapies alter tumor vasculature and thus close monitoring of changes in tumor blood flow could predict patient response. Currently, liver tumors are evaluated based on size with computed tomography (CT) or magnetic resonance imaging (MRI). However, contrast-enhanced ultrasound (CEUS) is perhaps better suited to track these blood flow changes than CT or MRI because it uses a vascular agent. With CEUS, blood flow-related parameters such as rise time (RT), mean transit time (MTT), peak intensity (PI) and area under the curve (AUC) – relating to blood velocity, volume, and distribution – can be extracted from CEUS video loops. Quantifying blood flow parameters allows for more sensitive and accurate evaluation of tumor response, but parameter reproducibility needs to be evaluated so that true changes in blood flow can be differentiated from measurement variation. The goal is to establish a standardized liver CEUS imaging and analysis protocol and evaluate the reproducibility of blood flow parameters. I analyzed CEUS scans collected with a standardized method from 80 patients with liver lesions using a MATLAB script to extract the parameters RT, MTT, PI, and AUC. I also performed the same analysis on images from an in-vitro study using the same methodology. I calculated the coefficient of variation (COV) of these parameters between scans to evaluate their reproducibility. The COVs indicate that the quantitative parameters are highly reproducible with agreement between in-vitro and clinical data. This shows that using the standardized methodology, reproducible blood flow parameters can be extracted from image loops and this technique can aid clinicians in the future to decide whether treatment is working.

Therapeutic ultrasound can induce biological effects that can be utilized for various clinical applications, and its non-invasiveness enables targeted treatments without harming tissue around the target. It can be applied in cancer treatments, where tumors can be primed with ultrasound to improve the delivery of chemotherapy, or even destroyed without the risks of surgery. Such treatment can be further enhanced by microbubbles, which are used clinically as a contrast agent in ultrasound imaging to visualize blood flow. Therapeutic ultrasound can generate microbubble activity known as cavitation that is capable of opening pores in cell membranes or disrupting blood supply to tumors, enabling more efficient drug uptake. My research goal has been to evaluate microbubble activity generated with therapeutic ultrasound and discover ways to optimize this treatment for drug delivery. To monitor microbubble activity during treatment, I use a technique known as passive cavitation detection (PCD) where one ultrasound device transmits sound directed at microbubbles, while the other “passive” device is listening for sound scattered off the microbubbles. I have been developing a PCD setup with a tissue-mimicking phantom that is physiologically similar to tumors for fast and reliable evaluation of ultrasound conditions for cavitation for use in cancer therapy. For this project, I align and control the PCD system with the LabView software, develop several phantoms that mimic cancer tissues for testing microbubble response to treatments, and analyze microbubble signals with a computation software MATLAB to evaluate cavitation activity. In addition to studying ultrasound cavitation, I am currently focusing on the fabrication of a tissue phantom with a cylindrical flow channel acting as a tumor blood vessel. The phantom allows for quick, repeatable experiments and evaluation of tumor vessels with different sizes. The careful study of cavitation activity will lead to more efficient cancer treatments with improved drug uptake.