THIS REPORT HAS BEEN WITHDRAWN

Arkosnato Neogy

EECS Department
University of California, Berkeley
Technical Report No. UCB/EECS-2013-50
May 8, 2013

Modern ultrasound imaging has superseded radiation-based (X-ray, CT, MRI) imaging modal- ities in applications that have high safety requirements. The low budget, both in cost and hospital real-estate, and real-time imaging possibilities in ultrasound have further increased the attractive- ness of this modality.

Ultrasound faces one primary drawback today - the lack of a range of contrast agents, that can suitably highlight selected tissues as opposed to the background. Since the acoustic impedance of most tissues in the human body is close to water, production of natural contrast cannot be expected. Therefore, standard contrasts agents such as microbubbles are used. These agents have their limi- tations. Specifically, the size of microbubbles makes it difficult for them to perfuse into capillary beds which is an important imaging target in various pathological conditions. In addition, cavi- tation based non-linear imaging physics for microbubbles increases risks of microbubbles being destroyed under ultrasound pressure unexpectedly, causing damage to surrounding healthy tissues.

To counter these drawbacks, our project focuses on proposing an alternative contrast agent for ultrasound imaging. Specifically, the target agent should be smaller in size, predominantly linear in its imaging response, while not losing out either in image quality (as measured by resolution, contrast and SNR) or in medical relevance (controlled sonication, drug delivery, controlled non- linear response). We discovered an unusual candidate to fit the above requirements - gas vesicles. Gas vesicles are sub-micron scale protein cages produced by photosynthetic bacteria and algae in aquatic environments to maintain buoyancy and receive optimal amount of sunlight.

Through the experiments described in this thesis, we establish the viability of gas vesicles as potential ultrasound contrast agents for future medical applications at a proof of concept level. A significant contributor to the success of using gas vesicles to produce imaging contrast is an auto- mated ultrasound apparatus developed indigenously for the purpose of experiments. The interplay of hardware functionality and flexible software control was designed so as to facilitate the choice of best-practice parameters to generate images of diagnostically acceptable quality. Using this ap- paratus, we establish the linearity of gas vesicles, and move on to demonstrate high quality images in terms of resolution, contrast and SNR. We extend our exploration further to advanced imaging techniques to ensure gas vesicles do not lose out on advantages offered by microbubble-like non-linear contrast agents. Specifically, we show that it is possible to use gas vesicles in non-linear regions for harmonic imaging, sonication and bio-functionalized imaging to establish its biological potential in future applications.

Advisor: Steven Conolly