Meeting Banner
Abstract #1664

Impact of Sampling Pattern on Combined Parallel Imaging Compressed Sensing Volumetric Knee MRI

Umar Tariq1, Peng Lai2, Michael Lustig3, Marcus T. Alley, Mian Zhang4, Garry E. Gold, Shreyas S. Vasanawala5

1Department of Radiology, Stanford University, Stanford, CA, United States; 2GE Healthcare, Menlo Park, CA, United States; 3University of Calfornia Berkeley, Berkeley, CA, United States; 4Statistics, Stanford University, Stanford, CA, United States; 5Radiology, Stanford University, Stanford University, CA, United States

We compare two compressed sensing sampling strategies to obtain thinner slices with higher acceleration, assessing image quality relative to gold-standard 2D images in the delineation of anatomic structures, generally evaluated in knee MRI.

Keywords

acceleration achieve achieved address adjusted alley analyses anatomic anterior approval arrow assent axial better blinded blurring bony capability cartilage channel clinical coil collateral combined combining comes comparable compressed confidence consist coronal cube defined delineated delineation density despite detrimental diagnostic domain entire environment equivalent error evaluate evaluated exams excluded exploits fashion fast fine five gaps generally gold good hypothesis impact improvements improves independently informed integrity john keep knee length lengthy ligament limited long makes matrix medial meniscus middle note null outer outstanding parallel park part pattern portions potential primarily primary problems prohibitively quality radiologists radiology randomized rank reader readers real received reconstructions recruited referred reformatted relatively representative requires resolution robustness sampling scale scholar score scores sensing sharpness short signed slice slices space sparsity spatial spin spirit statistical straighten strategies structure structures suboptimal suffers suggesting thick thin thinner though title torn train transferred true type uniform variable vista volumetric wide widely workstation years