In our stride to make sharper and clearer pictures using ultrasound, we have been exploring various possible strategies starting at the data acquisition level. Our tactical goal is to image physiological events that are impossible to detect using conventional ultrasound. We are especially interested in imaging dynamic events like blood flow and tissue motion using broad-view imaging paradigms that can acquire data at >1,000 fps. Various signal processing techniques such as least-squares estimation, adaptive beamforming, and eigen-component analysis are widely applied in our work.
As the new imaging techniques that we develop cannot be readily realized on existing scanners, we have been actively devising our own experimental hardware. Key initiatives include: 1) designing channel-domain Tx/Rx hardware, and 2) using GPUs and FPGAs as high-speed beamforming platforms. We are also actively developing imaging phantoms to help us systematically evaluate the efficacy of our new imaging methods
We have always been curious about how low-intensity ultrasound can deliver therapeutic impact to living cells. Of high interest to us is the interfacial impact of ultrasound at a single-cell level (for both mammalian and plant cells). Sonoporation (cavitation-induced membrane perforation) is one particular phenomenon that we have been studying.
We reckon that, for ultrasound to be applicable to non-ablation-type treatments, deep understanding of the bioeffects beyond the interfacial level is critical. We have been studying different subcellular bioeffects of sonoporation, including: 1) changes to cell-cycle dynamics; 2) disruption of cytoplasmic signaling pathways. We are expecting to gain new insight into how ultrasound-mediated stimulatory treatments and drug/gene delivery can be carried out effectively.