The use of parallel imaging for MRI assessment of knees in children and adolescents

Background Parallel imaging provides faster scanning at the cost of reduced signal-to-noise ratio (SNR) and increased artifacts. Objective To compare the diagnostic performance of two parallel MRI protocols (PPs) for assessment of pathologic knees using an 8-channel knee coil (reference standard, conventional protocol [CP]) and to characterize the SNR losses associated with parallel imaging. Materials and methods Two radiologists blindly interpreted 1.5 Tesla knee MRI images in 21 children (mean 13 years, range 9–18 years) with clinical indications for an MRI scan. Sagittal proton density, T2-W fat-saturated FSE, axial T2-W fat-saturated FSE, and coronal T1-W (NEX of 1,1,1) images were obtained with both CP and PP. Images were read for soft tissue and osteochondral findings. Results There was a 75% decrease in acquisition time using PP in comparison to CP. The CP and PP protocols fell within excellent or upper limits of substantial agreement: CP, kappa coefficient, 0.81 (95% CIs, 0.73–0.89); PP, 0.80–0.81 (0.73–0.89). The sensitivity of the two PPs was similar for assessment of soft (0.98–1.00) and osteochondral (0.89–0.94) tissues. Phantom data indicated an SNR of 1.67, 1.6, and 1.51 (axial, sagittal and coronal planes) between CP and PP scans. Conclusion Parallel MRI provides a reliable assessment for pediatric knees in a significantly reduced scan time without affecting the diagnostic performance of MRI.