Grant: Research Tools and Instruments

The primary goal of the requested infrastructure is to push the boundaries of machine learning solutions for ultrasound imaging, towards a generalizable learning ecosystem that can be deployed with data obtained at the edge in distributed point-of-care settings. In recent years, ultrasound imaging has experienced exponential growth in user numbers due to advances in miniaturization and the development of portable point-of-care ultrasound systems. However, addressing the challenge of the limited supply of highly trained ultrasound operators is crucial to expand access to underprivileged and rural areas. Machine learning integration with ultrasound scanners has shown significant promise in facilitating data acquisition and interpretation. However, existing machine learning-based approaches are limited in scope, as they are trained based on centralized, retrospective datasets, which raises questions about their generalizability when deployed across a range of ultrasound operators and patient distributions. Our goal is to disrupt the current learning paradigm for ultrasound imaging by designing new solutions that are specifically suited for a distributed system consisting of interconnected point-of-care ultrasound devices. Collaborating closely with a leading Canadian point-of-care ultrasound manufacturer, Clarius Mobile Health, we will investigate machine learning methodologies that can be deployed at the edge, while learning can be performed collectively within the network setting. Such learning approach will use continuous, federated learning to dynamically update a centralized model stored on a computing server to adjust to new data distribution on the edge, while optimizing the accuracy of the overall learning system for all point-of-care operators. As a research team with a demonstrated commitment to equity, diversity, and inclusion, we are well-positioned to further promote EDI across the entire research program, including the recruitment of HQP and their placement. HQP trained in this research program will gain the necessary skills to meet the pressing needs of the Canadian industry and academia. The demand for HQP with data analytics skills is growing substantially. These HQP will be well-prepared for employment in Canadian companies within the biomedical engineering sector and beyond.