Assessment of idiopathic inflammatory myopathy using muscle T2 mapping segmented by Deep Anatomical Federated Network (DAFNE)
Fengdan Wang1, Francesco Santini2, Jinxia Zhu3, Tom Hilbert4,5,6, Tobias Kober4,5,6, and Zhengyu Jin1
1Radiology, Peking Union Medical College Hospital, Beijing, China, 2University Hospital Basel, Basel, Switzerland, 3Siemens Healthcare Ltd., Beijing, China, 4Siemens Healthcare AG, Lausanne, Switzerland, 5Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland, 6École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland
Idiopathic Inflammatory Myopathy (IIM) is a group of immune-mediated myopathies, with high morbidity and mortality. Prior work showed that dedicated accelerated T2 mapping of bilateral thighs can be performed within 3 minutes at high resolution and can detect IIM-induced muscle inflammation. However, manual drawing methods to delineate regions of interest are subject to sampling errors. This study investigated the first utility of a fully automated deep-learning method for segmentation of T2 maps in 64 patients with IIM against healthy controls, with results confirming its feasibility, accuracy, and efficiency.
This abstract and the presentation materials are available to members only;
a login is required.