eduzhai > Applied Sciences > Engineering >

Extending LOUPE for K-space Under-sampling Pattern Optimization in Multi-coil MRI

  • king
  • (0) Download
  • 20210505
  • Save

... pages left unread,continue reading

Document pages: 11 pages

Abstract: The previously established LOUPE (Learning-based Optimization of theUnder-sampling Pattern) framework for optimizing the k-space sampling patternin MRI was extended in three folds: firstly, fully sampled multi-coil k-spacedata from the scanner, rather than simulated k-space data from magnitude MRimages in LOUPE, was retrospectively under-sampled to optimize theunder-sampling pattern of in-vivo k-space data; secondly, binary stochastick-space sampling, rather than approximate stochastic k-space sampling of LOUPEduring training, was applied together with a straight-through (ST) estimator toestimate the gradient of the threshold operation in a neural network; thirdly,modified unrolled optimization network, rather than modified U-Net in LOUPE,was used as the reconstruction network in order to reconstruct multi-coil dataproperly and reduce the dependency on training data. Experimental results showthat when dealing with the in-vivo k-space data, unrolled optimization networkwith binary under-sampling block and ST estimator had better reconstructionperformance compared to the ones with either U-Net reconstruction network orapproximate sampling pattern optimization network, and once trained, thelearned optimal sampling pattern worked better than the hand-crafted variabledensity sampling pattern when deployed with other conventional reconstructionmethods.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×