Meeting Banner
Abstract #1335

Sparse Representations for Compressed Sensing Acceleration of Fourier Velocity Encoded MRI

Gabriel L. S. L. Oliveira1, Joao L. A. Carvalho1

1Department of Electrical Engineering, University of Braslia, Braslia, DF, Brazil

Fourier velocity encoding (FVE) is useful in the assessment of vascular and valvular stenosis and intravascular wall shear stress, as it eliminates partial volume effects that may cause loss of diagnostic information in more conventional phase-contrast MRI. FVE shows great potential for compressed sensing acceleration, due to its high dimensionality and intrinsic sparseness in image domain. In this work, we investigate other sparse representantions for FVE data, using a five-dimensional (x,y,z,v,t) FVE dataset of the neck (focusing on carotid flow). Several combinations of separable transforms were evaluated. Two promising combinations of transforms are proposed.

Abstract PDF

Keywords

able acceleration acquisition adopted although among applications applied arise artifacts assessed assessment axial best better bifurcation blue bras cardiac carotid carotids cause channel cine clinical coefficients coil combination combinations combined compressed contrast cosine criterion cumulative curve curves dataset density descending diagnostic dimension dimensionality dimensions distributions domain efficiently electrical eliminates encoded encodes energy engineering evaluated evaluation excite experiment fastest finite five flow focusing form fulfilling gradients great green healthy heartbeats highlighted important included inset intrinsic investigate largest long loss neck nevertheless normalized outperforms partial patients potential primarily prohibitively promising proposed qualitative readouts realized reconstructed reconstruction refers representation representations resolution routine sample search searching selected sensing separable sets several shear slice slices sorted sparse sparseness sparsity spatial spiral step still stress successfully suggest system tang temporal toolbox towards transform transformed transforms uniform useful valvular variable various vascular velocity volume volunteer wall wavelet wavelets