Quantitative Ultrasound Imaging for Diagnosis and Intervention: A Machine Learning Approach

Abstract: In recent years, quantitative ultrasound has emerged as a promising technology to improve diagnosis, increase the precision of interventions, and automate the acquisition of data specially in point-of- care settings. Advancements in machine learning and computation power have contributed significantly to these developments in quantitative imaging.
In this talk, I present three main developments in our laboratory based on quantitative ultrasound:

• For prostate cancer diagnosis, I present our work based on the analysis of temporal ultrasound data. I compare the cancer detection and grading accuracy based on temporal ultrasound with multi-parametric MRI in a fusion biopsy study on more than 150 patients. Our results demonstrate that a combination of temporal ultrasound and MRI can accurately predict prostate histopathology determined by biopsy samples.
• For spine anesthesia and analgesia, I present the result of our work on combining anatomical models with ultrasound data to facilitate image interpretation in the spine region. Furthermore, I demonstrate that using advanced machine learning techniques, automation of image interpretation of the spine is feasible.
• For echocardiography, we have access to more than 200,000 patient records based on clinical data obtained over the last 10 years in Vancouver. I present the initial results of our research on 7,000 patients both in automatic analysis of data to improve measurement accuracy, and to assign an image quality factor to the data to facilitate image acquisition.

Bio: Purang Abolmaesumi is the Canada Research Chair in Biomedical Engineering, a Killam Research Prize recipient, and Professor in the Department of Electrical and Computer Engineering at the University of British Columbia, Vancouver, BC, Canada. He is internationally recognized and has received numerous awards for his pioneering developments in ultrasound image processing, image registration and image-guided interventions. He is the General Chair of the International Conference on Information Processing in Computer Assisted Intervention, 2014-2017, and has served on the program committees of the Medical Image Computing and Computing and Computer Assisted Intervention (MICCAI) and International Society for Optics and Photonics (SPIE) Medical Imaging. Dr. Abolmaesumi is an Associate Editor of the IEEE Transactions on Medical Imaging, and has served as an Associate Editor of the IEEE Transactions on Biomedical Engineering. His techniques for ultrasound segmentation have been adopted to develop a prostate brachytherapy planning interface that is now the standard-of- care in British Columbia and has been used to treat more than 2,000 patients.