The overarching goal of our research is to develop patient-centered and equitable engineering solutions to problems in radiology and surgery. Specifically, our research pairs ultrasound and photoacoustic imaging with advanced signal processing and machine learning techniques to improve cancer diagnosis and guide safer surgical procedures. Read about the most recent work below.
Coherence-based Imaging for Breast Ultrasound
Due to the heterogeneity of breast tissue, breast ultrasound images are often plagued with a noise artifact called acoustic clutter. With coherence-based beamforming techniques, I am studying how to remove this noise artifact and improve the diagnostic quality of breast ultrasound.
Deep Learning for Ultrasound Image Formation
The ultrasound beamforming process is often very time consuming. I leveraged the universal approximation properties of deep neural networks to speed up this beamforming process specifically for coherence-based beamforming techniques, while retaining the integrity and benefits of the original method.
Photoacoustic Imaging for Surgical Guidance
Photoacoustic imaging, which pairs pulsed laser light with ultrasound detection, has the ability to provide surgeons with both targeted and selective imaging. During my PhD, I focused on hysterectomy procedures, where a contrast agent can be used to differentiate the ureter from the uterine artery for safer hysterectomy procedures.