Detection and Characterization of the Synovium in Musculoskeletal Ultrasound

Arthritis is a chronic disease that severely decreases the quality of life and affects almost 4.6 million Canadians, costing $33 billion for the Canadian economy every year. Affected individuals experience pain and disability through an extended period of time. Rheumatoid arthritis (RA), a common form of arthritis, is an autoimmune disease characterized by the inflammation of the synovium, or synovial membrane, a connective tissue that provides a cushion between bones and tendons and muscle around a joint. Characterization of the synovium is crucial for the management of RA as well as a useful marker of treatment efficiency. Ultrasound imaging is an affordable non-invasive technique to assess the synovium for thickening. The examination of the images is often contingent on sonographers and clinicians. Therefore, the current research proposes to utilize deep learning to accurately and reliably assess the synovium on ultrasound images, and then implement the network in the software implementable in medical settings.

Faculty Supervisor:

Pascal Tyrrell

Student:

Dmitrii Paniukov

Partner:

16 Bit Inc.

Discipline:

Medicine

Sector:

Professional, scientific and technical services

University:

University of Toronto