Meeting Banner
Abstract #2373

MRiLab: Performing Fast 3D Parallel MRI Numerical Simulation on a Simple PC

Fang Liu1, Richard Kijowski2, Walter F. Block1, 3

1Department of Medical Physics, University of Wisconsin-Madison, Madison, WI, United States; 2Department of Radiology, University of Wisconsin-Madison, Madison, WI, United States; 3Department of Biomedical Engineering, University of Wisconsin-Madison, Madison, WI, United States

A simulation package named MRiLab is introduced here for performing efficient and flexible 3D numerical MRI simulation on simple PC. MRiLab features user-friendly interface and optimized programming-free modules that well suits for MR pulse sequence design and image reconstruction technique development. Associated tools are also provided for analyzing multiple spin evolution, MR signal and k-space formation, local SAR distribution, etc. With the simulation acceleration accompanied by GPU and multi-threaded CPU, MRiLab has proven to be a fast and effective parallel simulation framework with minimum computation hardware requirements.

Keywords

able accelerated accurate acquisition added adiabatic adjusting aimed alternative although analyzing architecture artifact aspects available axial banding behavior biomedical block bottom build caused channel circular cluster coding coil comparable composed comprehensive computation computer conducting contains control cost creating customized dedicated design designing desktop developing development diagram diffusion digital discrete distribution educational effective either employ environment equation evolution exchange expensive exploits facilitate fang fast feasible features field free function functional functions gives gradient graphical graphically great hardware highly holds ideas improvement includes incorporation inhomogeneity instantly interactive interface kernel library like limited local magnet many module modules named note numerical ordinary package panel parallel parallelized phantom physics pool potentials practical precession profile programming providing pulse real reconstruction requiring resolution resultant running saved seamlessly selective setup side simple simplifies simulated simulating simulation slice solution solving speed spin steady steps stocker storing studies system threaded tool tools transition translates tree typically understanding unit variation various