Meeting Banner
Abstract #3793

Fast Compressed Sensing Reconstruction Using a Direct Fourier-Wavelet Transform

KyungHyun Sung1, Brian Andrew Hargreaves2

1Radiological Sciences, UCLA, Los Angeles, CA, United States; 2Radiology, Stanford University, Stanford, CA, United States

High computational complexity is one major issue for compressed sensing (CS) reconstruction. We present a new way to reduce the computational complexity for the CS reconstruction by directly transforming between k-space and wavelet domains. This replaces FIR filtering in the image domain with a multiplication in k-space and can reduce computational complexity. This efficient computation can benefit almost all CS methods that exploit the wavelet sparsity, and we have shown the actual CS reconstruction time can be reduced by 46% on MATLAB.

Keywords

accentuate actual adjustments allows almost alternatives amount anticipate application applied apply approaches approximate avoid backward benefit bottleneck boundary bounds breast challenging code comm complexity compressed computation computational computationally consisting convex core deviation diagram direct directly disabled domain domains dual efficient employed enabled equipped error evaluated even exactly executable exists exploit extensively extremely fast faster filter filtering filters finite fixed forward frames function good identical illustrates illustration implement implementation implemented importantly improvements impulse includes inverse issues iterations iterative known load maintain major math matrix measured memory message minimization minimize multiplication multiplications noise note often operation operations optimization pass passing primarily problems process processes propose proposed pure quality radiological radiology reconstructing reconstruction reconstructions reduce reduced remains repetitions replaced replaces representation response retrospectively samples sciences secondly sensing serial since soft solve solving space sparsity spectral still studied subtle sung temporal thresholding transform transforming transforms variation vector volume wavelet weightings