Knee menisci segmentation and relaxometry of 3D ultrashort echo time cones MR imaging using attention U‐Net with transfer learning

• Published in: Magnetic resonance in medicine Vol. 83; no. 3; pp. 1109 - 1122
• Main Authors: Byra, Michal, Wu, Mei, Zhang, Xiaodong, Jang, Hyungseok, Ma, Ya‐Jun, Chang, Eric Y., Shah, Sameer, Du, Jiang
• Format: Journal Article
• Language: English
• Published: United States Wiley Subscription Services, Inc 01.03.2020
• Subjects: Adult; Aged; Aged, 80 and over; Algorithms; Artificial neural networks; Biomedical materials; Cones; Correlation coefficients; Data collection; Deep Learning; Female; Humans; Image processing; Image Processing, Computer-Assisted - methods; Image segmentation; Imaging, Three-Dimensional; Knee; Linear Models; Machine learning; Magnetic Resonance Imaging; Male; Mathematical models; menisci; Menisci, Tibial - diagnostic imaging; Meniscus; Middle Aged; Neural networks; Neural Networks, Computer; Observer Variation; Osteoarthritis; Parameter estimation; Pattern Recognition, Automated; quantitative MR; Radiology; Reproducibility of Results; segmentation; Transfer learning; Young Adult; deep learning; segmentation; quantitative MR; menisci; osteoarthritis

Purpose To develop a deep learning‐based method for knee menisci segmentation in 3D ultrashort echo time (UTE) cones MR imaging, and to automatically determine MR relaxation times, namely the T1, T1ρ, and T2∗ parameters, which can be used to assess knee osteoarthritis (OA). Methods Whole knee joint...