Developing Imaging Methods and Processing Techniques for Super-resolution Ultrasound Microvascular Applications

Super-resolution (SR) ultrasound provides imaging resolutions beyond the typical diffraction-limited resolutions of conventional ultrasound, enabling superficial and non-superficial microvasculature on the order of tens of microns to be acoustically resolved. This is also known as ultrasound localization microscopy (ULM). In ULM, localizing and accumulating the positions of separable sub-wavelength microbubble contrast agents (MCAs) allows for the elimination of the typical blurred appearance associated with the physics-determined imaging point-spread-function, allowing for higher (super-) resolution results. SR ultrasound imaging offers a non-invasive way to image vascular morphologies, which could be of particular benefit for patient pathologies where the characterization or monitoring of (micro)vasculature is diagnostically useful. The research in this dissertation explores three distinct, novel approaches for SR ultrasound imaging, each contending with a different aspect of the imaging or processing pipeline and therefore a different challenge towards clinical translation. First, I explore the use of phase-change contrast agents (PCCAs) as an alternative to conventional MCA contrast agents by using the unique acoustic signature of a PCCA vaporizing as a localization target for ULM imaging. Dual-frequency imaging approaches could offer this new approach high contrast-to-tissue ratios (CTRs) and I demonstrate the SR capabilities in vitro in a flow-phantom. Second, I examine the use of correlation-based methods as an alternative spatio-temporal filtering step versus the SVD. I demonstrate a novel application of a highly sensitive, correlation-based motion-detection algorithm as an approach for MCA signal isolation. I use In silico experiments to characterize the approach's abilities and in vivo transcranial rat brain data to demonstrate the SR capabilities of this approach and compare it to the SVD approach. And third, I demonstrate the use of a robot arm to collect angled, co-registered, compounded SR images to offer a method of increasing imaging field of view. For transcranial imaging where ultrasound pressure through the skull is critical (which incident angle with respect to the skull geometry contributes to), unfocused imaging approaches fail to appropriately insonify the whole brain and thus fail to resolve certain brain vasculature. Therefore, in vivo transcranial volumetric imaging is used to demonstrate the advantages of this robot-arm compounding technique.