Jan Fritz1, Benjamin Fritz2, Gaurav K Thawait3, Esther Raithel4, Wesley D Gilson5, Mathias Nittka4, Michael A Mont6. 1. Russell H. Morgan Department of Radiology and Radiological Science, Section of Musculoskeletal Radiology, Johns Hopkins University School of Medicine, 601 N Caroline Street, JHOC 3140A, Baltimore, MD, 21287, USA. jfritz9@jhmi.edu. 2. Department of Radiology, University of Freiburg, Robert-Koch-Straße 3, 79106, Freiburg im Breisgau, Germany. 3. Russell H. Morgan Department of Radiology and Radiological Science, Section of Musculoskeletal Radiology, Johns Hopkins University School of Medicine, 601 N Caroline Street, JHOC 3140A, Baltimore, MD, 21287, USA. 4. Siemens Healthcare GmbH, Allee am Roethelheimpark 2, 91052, Erlangen, Germany. 5. Siemens Healthcare USA, Inc., 600 N. Wolfe Street, Park 320, Baltimore, MD, 21287, USA. 6. Department of Orthopedic Surgery, Cleveland Clinic Foundation, Cleveland, OH, USA.
Abstract
OBJECTIVE: Compressed sensing (CS) acceleration has been theorized for slice encoding for metal artifact correction (SEMAC), but has not been shown to be feasible. Therefore, we tested the hypothesis that CS-SEMAC is feasible for MRI of metal-on-metal hip resurfacing implants. MATERIALS AND METHODS: Following prospective institutional review board approval, 22 subjects with metal-on-metal hip resurfacing implants underwent 1.5 T MRI. We compared CS-SEMAC prototype, high-bandwidth TSE, and SEMAC sequences with acquisition times of 4-5, 4-5 and 10-12 min, respectively. Outcome measures included bone-implant interfaces, image quality, periprosthetic structures, artifact size, and signal- and contrast-to-noise ratios (SNR and CNR). Using Friedman, repeated measures analysis of variances, and Cohen's weighted kappa tests, Bonferroni-corrected p-values of 0.005 and less were considered statistically significant. RESULTS: There was no statistical difference of outcomes measures of SEMAC and CS-SEMAC images. Visibility of implant-bone interfaces and pseudocapsule as well as fat suppression and metal reduction were "adequate" to "good" on CS-SEMAC and "non-diagnostic" to "adequate" on high-BW TSE (p < 0.001, respectively). SEMAC and CS-SEMAC showed mild blur and ripple artifacts. The metal artifact size was 63 % larger for high-BW TSE as compared to SEMAC and CS-SEMAC (p < 0.0001, respectively). CNRs were sufficiently high and statistically similar, with the exception of CNR of fluid and muscle and CNR of fluid and tendon, which were higher on intermediate-weighted high-BW TSE (p < 0.005, respectively). CONCLUSION: Compressed sensing acceleration enables the time-neutral use of SEMAC for MRI of metal-on-metal hip resurfacing implants when compared to high-BW TSE and image quality similar to conventional SEMAC.
OBJECTIVE: Compressed sensing (CS) acceleration has been theorized for slice encoding for metal artifact correction (SEMAC), but has not been shown to be feasible. Therefore, we tested the hypothesis that CS-SEMAC is feasible for MRI of metal-on-metal hip resurfacing implants. MATERIALS AND METHODS: Following prospective institutional review board approval, 22 subjects with metal-on-metal hip resurfacing implants underwent 1.5 T MRI. We compared CS-SEMAC prototype, high-bandwidth TSE, and SEMAC sequences with acquisition times of 4-5, 4-5 and 10-12 min, respectively. Outcome measures included bone-implant interfaces, image quality, periprosthetic structures, artifact size, and signal- and contrast-to-noise ratios (SNR and CNR). Using Friedman, repeated measures analysis of variances, and Cohen's weighted kappa tests, Bonferroni-corrected p-values of 0.005 and less were considered statistically significant. RESULTS: There was no statistical difference of outcomes measures of SEMAC and CS-SEMAC images. Visibility of implant-bone interfaces and pseudocapsule as well as fat suppression and metal reduction were "adequate" to "good" on CS-SEMAC and "non-diagnostic" to "adequate" on high-BW TSE (p < 0.001, respectively). SEMAC and CS-SEMAC showed mild blur and ripple artifacts. The metal artifact size was 63 % larger for high-BW TSE as compared to SEMAC and CS-SEMAC (p < 0.0001, respectively). CNRs were sufficiently high and statistically similar, with the exception of CNR of fluid and muscle and CNR of fluid and tendon, which were higher on intermediate-weighted high-BW TSE (p < 0.005, respectively). CONCLUSION: Compressed sensing acceleration enables the time-neutral use of SEMAC for MRI of metal-on-metal hip resurfacing implants when compared to high-BW TSE and image quality similar to conventional SEMAC.
Authors: K M Koch; A C Brau; W Chen; G E Gold; B A Hargreaves; M Koff; G C McKinnon; H G Potter; K F King Journal: Magn Reson Med Date: 2011-01 Impact factor: 4.668
Authors: Gunilla M Müller; Björn Lundin; Thord von Schewelov; Markus F Müller; Olle Ekberg; Sven Månsson Journal: Skeletal Radiol Date: 2014-12-05 Impact factor: 2.199
Authors: Cesar de Cesar Netto; Lucas F Fonseca; Benjamin Fritz; Steven E Stern; Esther Raithel; Mathias Nittka; Lew C Schon; Jan Fritz Journal: Eur Radiol Date: 2017-12-07 Impact factor: 5.315
Authors: Tim Fischer; Christoph Stern; Benjamin Fritz; Patrick O Zingg; Christian W A Pfirrmann; Reto Sutter Journal: Skeletal Radiol Date: 2020-06-25 Impact factor: 2.199