Isotropic MRI of the Knee at 3.0T with the 3D-TSE-sequence Sampling Perfection with Application Optimized Contrasts Using Different Flip Angle Evolutions (SPACE) – First Clinical Experiences

M. Notohamiprodjo; A. Horng; M. Pietschmann; W. Horger; K. A. Herrmann; M. F. Reiser; C. Glaser

doi:10.1055/s-0028-1124045

RöFo - Fortschritte auf dem Gebiet der Röntgenstrahlen und der bildgebenden Verfahren, Inhaltsverzeichnis

Isotropic MRI of the Knee at 3.0T with the 3D-TSE-sequence Sampling Perfection with Application Optimized Contrasts Using Different Flip Angle Evolutions (SPACE) – First Clinical Experiences

M Notohamiprodjo ¹, A Horng ¹, M Pietschmann ², W Horger ³, KA Herrmann ¹, MF Reiser ¹, C Glaser ¹

¹Department of Clinical Radiology, University of Munich Hospitals, Grosshadern Campus, Munich, Germany
²Department of Orthopedics, University of Munich Hospitals, Grosshadern Campus, Munich, Germany
³Siemens Medical Solutions, Erlangen, Germany

Abstract

Purpose: To evaluate a fat-saturated moderately T2w-3D-TSE-sequence(SPACE) for high-resolution isotropic knee-MRI at 3T.

Materials and Methods: For technical and anatomical detail assessment sagittally oriented SPACE-datasets (TR 1200ms/TE 30ms/0.5³mm³) and standard 2D-TSE-sequences in three planes (TR 3200ms/TE 30ms/0.4×0.4×3mm) were acquired from the dominant knee of 10 healthy volunteers on a 3T-scanner. 0.5, 1 and 2mm thick multiplanar SPACE-reconstructions (MPR) were performed in the three major anatomic planes. For lesion detection assessment, 60 patients with cartilage and meniscus pathologies (18 with arthroscopic correlation) were examined with the same techniques. Visual assessment was independently performed by 2 readers. Statistical analysis was performed with paired t-tests and Weighted-Kappa for interreader-correlation.

Results: Acquisition time was 10'35'' for SPACE and 12'34'' for 2D-TSE. Signal-to-noise- (SNR) and contrast-to-noise-ratios (CNR) of the original isotropic SPACE-sequence were low, SNR and CNR of 1mm- and 2mm-MPR comparable to 2D-TSE. 1mm-MPR provided the best trade-off between image-quality and anatomical information. Visualization of small ligamentous structures was superior for isotropic- and 1mm-MPR compared to 2D-TSE and 2mm-MPR. Interreader-correlation for anatomical detail visualization was good (r=0.73–0.77) for SPACE and significantly higher than for 2D-TSE (r=0.58).

Patients: The differences in detection and visualization of meniscus and cartilage lesions were not significant. However, SPACE was superior for meniscus lesions, whereas 2D-TSE was slightly superior for cartilage lesions. Interrreader correlation was good for meniscus lesions (r=0.67–0.73) and moderate for cartilage lesions (r=0.51–0.54) for both sequences.

Conclusion: Isotropic SPACE-datasets with retrospective 3D-MPR for knee-MRI at 3T are time efficient and provide good anatomical detail. Meniscus and cartilage lesion visualization equals current 2D-sequences.