(Sunday, 7/29/2018) 4:00 PM - 4:55 PM
Room: Room 207
Purpose: The linearized geometric solution (LGS) offers a novel remedy to the problematic off-resonance artifacts that plague balanced steady state free precession (bSSFP) MRI; the recent discovery that the LGS also mitigates MRI motion artifacts is a watershed moment that could allow it to be employed beyond its standard application in cardiac imaging. The goal here is to investigate the extent of MRI motion artifact correction that the LGS offers.
Methods: Four ∆θ = 0°, 90°, 180°, and 270° phase-cycled bSSFP datasets were simulated with flip angle = 70°, repetition time = 4.2ms, spatially varying T1 = 200 → 3000 ms, T2 = 40 → 3000 ms, and off-resonant accumulated phase θ = –π to π. Basic shapes were overlaid via a binary mask, and motion was simulated by translating the shapes of the ∆θ = 0° dataset stepwise with an overall oscillation frequency and amplitude. Motion frames were Fourier transformed, transformed data obliquely sampled, oblique k-space inverse-Fourier transformed, and zero-mean gaussian noise added to yield motion-corrupted images. The LGS and complex sum (CS) of the four phase-cycled images (1 corrupted by motion) were computed, the total relative error (TRE) with respect to a non-motion dataset tabulated, and the TRE plotted as a function of motion frequency, motion amplitude, noise level and T1/T2 relaxation ratio.
Results: The LGS corrects off-resonance-induced dark banding and also mitigates motion artifact better than CS in most motion scenarios. The only exceptions are when motion has low frequencies and high amplitude, and at very high T1/T2 or noise values.
Conclusion: The LGS reconstruction of bSSFP images mitigates motion artifacts and eliminates off-resonance-induced dark banding. This removes major obstacles to exploiting bSSFP’s high efficiency, and will potentially increase its clinical applications.
Keywords
Not Applicable / None Entered.
Taxonomy
IM- MRI : Image Reconstruction
Contact Email
