Gregory Nevsky MD, Presenter: Nothing to Disclose

Bei Zhang, Abstract Co-Author: Nothing to Disclose

Graham Wiggins, Abstract Co-Author: Nothing to Disclose

Soterios Gyftopoulos MD, Abstract Co-Author: Nothing to Disclose

Michael Paul Recht MD, Abstract Co-Author: Nothing to Disclose

Ravinder Regatte PhD, Abstract Co-Author: Nothing to Disclose

Gregory Chang MD, Abstract Co-Author: Nothing to Disclose

The performance of ultra high field 7T MRI for clinical knee imaging has not been studied. The goal of this study was to compare signal-to-noise ratio (SNR), image quality, and radiologist preferences for 3T and 7T clinical knee MRI.  

This study had IRB approval, and informed consent was obtained from all subjects. We recruited eight patients (mean age 35.4, 7 male, 1 female) with nine ACL reconstructions. We scanned each reconstructed knee using a 15 channel knee coil at 3T and a 28 channel knee coil at 7T on the same day. Our MRI protocol was coronal and sagittal proton density (PD) with fat saturation (FS) and sagittal PD without FS (TR/TE 3270/26, FOV 160 mm, matrix 732 x 896, slice thickness 3 mm) and axial PD with FS (TR/TE 1800/26 , FOV 131x160 mm, matrix 896 x 732, slice thickness 3 mm). Parallel imaging was utilized with an acceleration factor of 2. Quantitative SNR maps were calculated using Kellman’s method (Magn Reson Med 2005; 54:1439-1447). 3 musculoskeletal radiologists (20, 2, 2 years experience) were shown de-annotated 3T and 7T images for each patient in random order and scored image quality based on the ability to visualize ligaments (L), menisci (M), cartilage (C), and bone marrow edema (BME) on a 5 point scale (5=excellent, 1=not visualized). The radiologists also indicated which set of de-identified images they preferred to read. We performed two-tailed t-tests to determine whether differences in image quality were significant. We calculated binomial probability to determine whether radiologists’ preferences were significant.  

Quantitative SNR maps (Figure) show approximately 2.3 fold higher SNR for 7T compared to 3T. Mean image quality for 3T/7T was 5±0/5±0 (L), 4.7±0.6/5±0 (C), 5±0/5±0 (M), and 5.0±0/4.0±0 (BME) (p≥0.42 for all). Overall, radiologists preferred 7T to 3T images in 20/27 cases (p<0.02). 

This is the first study to compare 3T and 7T MRI for clinical knee imaging (ACL reconstructed patients). 7T images have greater than two-fold the SNR and were slightly preferred by MSK radiologists compared to 3T images, which were still excellent in quality. Protocol optimization to account for altered relaxation rates will likely improve the visualization of bone marrow edema at 7T.    

Radiologists slightly prefer 7T over 3T. Since 3T image quality is already excellent, greater SNR at 7T could be used to speed imaging or permit time for biochemical MRI.

Nevsky, G, Zhang, B, Wiggins, G, Gyftopoulos, S, Recht, M, Regatte, R, Chang, G, Clinical Knee MRI: Is 7T Better than 3T?.  Radiological Society of North America 2012 Scientific Assembly and Annual Meeting, November 25 - November 30, 2012 ,Chicago IL. http://archive.rsna.org/2012/12043690.html