Modification of Ultrasound Contrast Agent for Optimal Imaging

Jeffrey Rowan

Supervised by Marvin Doyley

601 Computer Studies Building

Modern medical imaging has brought us from a world requiring exploratory surgery to the one where diagnosis can occur with little to no effect on the surrounding tissue. While it can be hard to imagine where to go from here, new advances in ultrasonic imaging can guide us towards more complete diagnostic capabilities. Unlike magnetic resonance imaging (MRI) or computed tomography (CT), ultrasonic imaging is a low cost, non-ionizing means to image tissue and can also be used for therapy in a wide variety of applications (Stride, 2015). Ultrasound will likely never replace these technologies, but it may reduce the financial and clinical burdens, providing more data, at lower costs, reducing exposure to ionizing radiation. One of the main disadvantages of current ultrasound techniques is the trade-off between penetration and resolution, but thanks to an observation made over fifty years ago, the gap may soon be bridged. The injection of air during a routine cardiac exam gave researchers an invaluable tool for improving backscattered ultrasound signals, the gas bubble.

With time, this idea has evolved. Inherently unstable gases have been exchanged for heavier ones leading to less diffusion. Protein, lipid, and polymer shells have been added to further stabilize them, increasing lifetime from seconds to months, with added benefits from the physics of encapsulation (Helfield, 2019; Stride, 2015). Size has been restricted to less than 10 microns, ensuring minimal filtration by the lungs. A modified wave equation that had been previously used to discover the sounds of rain droplets now applies to the optimization of insonication via the linearly approximated resonant frequency (Marmottant et al., 2005). The culmination of the last 50 years of research has identified two main factors of resonant frequency are initial size and surface tension, although the latter is less noticeable for larger bubbles and has not been extensively investigated for (>3 micron). The uses of these bubbles in the clinical setting is near, as reports of optimizing bubble size for super-resolution imaging, super-harmonic imaging, and subharmonic pressure estimation (Lin et al., 2017; Kierski et al., 2019; Gupta et al., 2019) have already seen improvements. However, these papers examine modern formulations of bubbles, with any bubble manipulation focused strictly on sizing. The current proposal aims to address both size and shell properties via the use of porous silicon nitride membranes.

Following successful filtration/separation of a commercial contrast agent using these membranes in a centrifuge device, the acoustic response will be examined and compared to preexisting filtered agents. Of particular interest is the subharmonic response, as it is thresholded and dependent on the initial surface tension of a gas bubble and its elasticity (Frinking et al., 2010). To confirm any difference in acoustic response is based on shell properties and sizing, a micro-pipetting technique previously applied to neutrophils (white blood cells) is used to measure the initial surface tension of bubbles directly, for lipid and protein shelled. It has the added benefit of approximating the deformation of bubbles encountering the membrane, and behavior leading to subharmonic response such as nucleation, lipid packing, and gas diffusion. A tangential flow separation method is applied to protein shelled agents as they are too heavy for centrifugation without destruction.