Meeting Banner
Abstract #1416

Computer Aided Detection of Synovial Abnormalities Near Total Hip Replacements on 3D-MSI MRI using Deep Neural Networks

Kevin Koch1, Ruben Stern2,3, Robin Karr1, Andrew Nencka1, Matthew F Koff2, and Hollis Potter2

1Radiology, Medical College of Wisconsin, Milwaukee, WI, United States, 2Magnetic Resonance Imaging, Hospital for Special Surgery, New York, NY, United States, 3Center for Data Science, New York University, New York, NY, United States

3D-MSI increases the visibility of a large number of important pathologies commonly found near implanted orthopaedic hardware, including: host-mediated adverse local tissue reactions, infection, osteolysis, and osteonecrosis. MRI identification of these pathologies aids in planning for surgical revision and has been shown capable of predicting tissue destruction in symptomatic hip replacements. Identification of these features is difficult, even for the interpreting physicians with substantial specialized training and experience . To address this current challenge, a deep-learning based pattern classification approach using 3D-MSI MRI is proposed and demonstrated to predict patterns of adverse synovial responses near hip replacements.

This abstract and the presentation materials are available to members only; a login is required.

Join Here