Quantitative susceptibility mapping using deep neural network: QSMnet


    Jaeyeon Yoona, 1, Enhao Gongb, c, 1, Itthi Chatnuntawechd, Berkin Bilgice, Jingu Leea, Woojin Jung a, Jingyu Koa, Hosan Junga, Kawin Setsompope, Greg Zaharchuk c, Eung Yeop Kimf, John Paulyb & Jongho Leea


    a. Laboratory for Imaging Science and Technology, Department of Electrical and Computer Engineering, Seoul National University, Seoul, South Korea
    b. Department of Electrical Engineering, Stanford University, Stanford, CA, USA
    c. Department of Radiology, Stanford University, Stanford, CA, USA
    d. National Nanotechnology Center, Pathum Thani, Thailand
    e. Department of Radiology, Harvard Medical School, Boston, MA, USA
    f. Department of Radiology, Gil Medical Center, Gachon University College of Medicine, Incheon, South Korea
    1. Jaeyeon Yoon and Enaho Gong equally contributed to the paper.


    Neurolmage 179 (2018)


        • New QSM reconstruction, QSMnet, is developed using a deep neural network.

        • QSMnet generates a highly accurate QSM map close to a gold standard (COSMOS) map.

        • Processing time of QSMnet is only a few seconds, achieving real-time processing.

        • In patients, QSMnet delivers similar lesion contrasts to conventional QSM.