Meeting Banner
Abstract #2558

Highly Accelerated SEMAC Metal Implant Imaging Using Joint Compressed Sensing and Parallel Imaging

Mathias Nittka1, Ricardo Otazo2, Leon D. Rybak2, Kai Tobias Block3, Christian Geppert4, Daniel K. Sodickson2, Michael P. Recht2

1Siemens AG, Erlangen, Germany; 2Department of Radiology, New York University School of Medicine, New York, United States; 3Department of Radiology, NYU Langone Medical Center, New York, United States; 4Siemens Medical Systems, New York, United States

A highly accelerated implementation of SEMAC for metal implant imaging is presented, which aims to achieve efficient metal artifact redcution at clinically tolerable scan times. An approach of joint compressed sensing and parallel imaging is used, kz-kz undersampling is based on a Poisson-disk pattern with fully sampled k-Space for autocalibration. Experiments on a cadaver knee with a joint replacement were carried out both with a 4-channel flex coil and a 15 channel TX/RX coil. First results show good metal artifact reduction without significant loss in image quality at a total acceleration factor of 6.9.

Keywords

accelerate accelerated accelerating acceleration accelerations acceptable achieve achieved acquisition acquisitions adaptive alloy alloys application artifact artifacts auto bandwidth block cadaver calibration candidate carried causing channel clinical clinically coil coils combine combining compressed conducted consequently containing contrast coronal coverage currently customized dataset denoted density describe desirable dimension dimensions disk distortions domain enable encoding enforcing equivalent exams experimental explore feasible field flex fold full fully future good highly human identical implant implants implementation important in vivo includes incorporates inherently initial iterative joint jointly knee latter like limited limits loss matrix medical medicine metal moderate motion near noise note novel orientations orthopedic parallel particularly patient pattern position potential previous problem quality radiology random readout recent reconstructed reconstruction reduce reducing reduction refinements replacement replacements resolution respect resultant risk sampled sampling scanner scheme schemes school sensing sensitivities significantly slice slices slightly soft solutions space sparse sparsity steps strengths studies suggest target thresholding tolerated transform turbo utility variable white