Medical Engineering Defense, David Garrett

https://caltech.zoom.us/j/81103249075

Modern healthcare relies on imaging modalities that visualize internal anatomy and pathology. While X-ray computed tomography (CT) and magnetic resonance imaging (MRI) provide clinically useful imaging across many applications, they face significant barriers to more frequent use: ionizing radiation limits repeated CT scanning, and MRI's high cost and long acquisition times create access disparities. Conventional handheld ultrasonography enables rapid, low-cost imaging but remains limited by narrow fields of view, operator dependence, and challenging image interpretation. Photoacoustic tomography has emerged as a promising alternative that combines optical absorption contrast with acoustic detection, offering molecular specificity without ionizing radiation. However, conventional photoacoustic imaging remains limited to depths of several centimeters, inhibiting applications in deep-tissue imaging like gastrointestinal or whole-body assessment.

In this thesis, we develop three approaches to extract clinically relevant information at human scales: ultrasound, thermoacoustic, and photoacoustic tomography. All three modalities leverage a custom 512-element, 60 cm diameter receiver array designed to detect acoustic signals across human-scale geometries. We validate these approaches through in vivo imaging, ex vivo tissue experiments, and phantom studies. First, we demonstrate ultrasound tomography of full human cross-sections in the abdomen and lower extremities, reconstructing backscatter contrast alongside quantitative maps of the speed of sound and attenuation coefficient. We show that ultrasound tomography enables visualization of features such as the liver, vasculature, muscle, and subcutaneous adipose across entire 2D human cross-sections. Second, we develop a thermoacoustic approach to guiding microwave ablation procedures. By modulating the microwave signal delivered through the probe, we record the generated thermoacoustic signals and use them to model the thermal dynamics during ablation. We show that this approach yields more accurate estimates of ablation zone geometry than standard look-up tables, which could allow for more precise ablation therapy. Third, we develop a method to extend the imaging depth of photoacoustic tomography using a wireless, ingestible capsule-based optical source. We demonstrate imaging depth up to 12 cm, which could open the door to photoacoustic imaging of regions like the gastrointestinal tract. Together, these approaches aim to expand the range of safe, informative, and accessible imaging modalities available to patients and clinicians.

Advisor: Professor Lihong Wang