RSNA 2008 

Abstract Archives of the RSNA, 2008


SSG13-06

Magnetic Resonance Imaging of Cortical Bone Using a 3D Ultrashort TE Pulse Sequence

Scientific Papers

Presented on December 2, 2008
Presented as part of SSG13: Musculoskeletal (Osteoporosis)

 Research and Education Foundation Support

Participants

Jiang Du, Presenter: GE Healthcare
Mark Bydder PhD, Abstract Co-Author: Research grant, General Electric Company
Christine B. Chung MD, Abstract Co-Author: Nothing to Disclose
Graeme Mervyn Bydder MBChB, Abstract Co-Author: Researcher, General Electric Company

PURPOSE

Cortical bone has an extremely short T2 and shows zero signal intensity when examined with clinical MR pulse sequences. We have developed a three dimensional ultrashort TE (3D UTE) pulse sequence which has TE reduced to 8 μs through a combination of a hard radiofrequency (RF) pulse excitation, radial ramp sampling and fast transmit/receive switching. The objectives of this study were to image cortical bone from both cadaveric samples and healthy volunteers using 3D UTE sequence on a 3T clinical scanner.

METHOD AND MATERIALS

The 3D UTE sequence employs a short duration (40 μs) hard RF pulse for signal excitation, which is them spatially encoded using 3D radial projection reconstruction. Long T2 signals were suppressed using three approaches: 1) a long duration 900 pulse followed by gradient dephasing, 2) an adiabatic inversion pulse and signal nulling, and 3) dual echo acquisition followed by echo subtraction. Cortical bone from eight samples and six healthy volunteers was imaged. The acquisition parameters included an isotropic FOV of 10 cm for cadaveric samples and 16 cm for human volunteers, TR of 23 to 94 ms, TI of 38 ms (for the inversion approach only), readout of 256, projections of 20000 to 60000 in a total scan time of 15 to 24 minutes. The projection data was re-gridded onto a 256x256x256 matrix and was followed by 3D Fourier transformation to produce the final UTE images.

RESULTS

High resolution and contrast images were achieved for all 8 bone samples and 6 volunteers with an isotropic spatial resolution of 0.39x0.39x0.39 mm3 for samples and 0.62x0.62x0.62 mm3 for volunteers. Of the three long T2 suppression approaches, the long T2 saturation approach provides the highest SNR of 25.1 which was significantly higher than 15.3 for the adiabatic inversion approach. The latter approach provides more uniform suppression of long T2 muscle and fat signals, and resulted in a 12% CNR increase. The combination with dual echo subtraction technique further reduced the residual long T2 signals.

CONCLUSION

The 3D UTE sequence provides high signal and image contrast for imaging cortical bone, which is undetectable with clinical MR pulse sequences. Long T2 saturation combined with dual echo subtraction is preferred for clinical use of the 3D UTE sequence.

CLINICAL RELEVANCE/APPLICATION

The 3D UTE sequence provides direct imaging of previously undetectable cortical bone and may find applications for bone quality characterization.

Cite This Abstract

Du, J, Bydder, M, Chung, C, Bydder, G, Magnetic Resonance Imaging of Cortical Bone Using a 3D Ultrashort TE Pulse Sequence.  Radiological Society of North America 2008 Scientific Assembly and Annual Meeting, February 18 - February 20, 2008 ,Chicago IL. http://archive.rsna.org/2008/6013689.html