![]() Here I will outline a typical use of topup and eddy (they are really intended to be used together) on a "typical" data set suited for use with eddy. The need to pass more information to eddy results in a more complicated command line. Some of the options in eddy, like for example the outlier detection and the slice-to-volume movement model, needs to know which slices were acquired together (in the case of multi-band) and the order in which this occurred. Hence we need to inform eddy about how each volume was acquired. An example of this would be acquisitions with different polarity of the phase-encoding. This means that eddy needs to be informed of the diffusion direction/weighting that was used for each volume.Įddy can utilise the information from different acquisitions that modulate how off-resonance translates into distortions. Unlike eddy_correct, eddy attempts to model the diffusion signal. This means we need to "inform" eddy of the results from topup (which is used to calculate the susceptibility distortions). The reasons for this areĮddy attempts to combine the correction for susceptibility and eddy currents/movements so that there is only one single resampling. Running eddy is a little bit more complicated than running for example its predecessor eddy_correct. ![]() I warmly recommend investing in a couple of CUDA cards. This is because the slow speed makes it almost impossible to test the more time-consuming options thoroughly. The eddy_cuda version is potentially much faster than eddy_openmp and not all new features will be available for the OpenMP version. There is no longer an executable named eddy and when I refer to the eddy-command in the rest of this users guide it is implied that this is either eddy_openmp or eddy_cuda. Your choice of CUDA version will depend on the CUDA version installed on your system. To create eddy_cuda if you do not have it on your system, simply link it to either eddy_cuda8.0, or eddy_cuda9.1. This is a one-time setup step per FSL installation. Therefore, some users may need to create the eddy_cuda sym-link on their system, and for their desired CUDA version. For all other users, it is not trivial to automatically detect where your system's CUDA installation is located without searching the entire file system. ![]() Within FMRIB, eddy_cuda will point to the newest CUDA installed on the cluster. This has been done in two ways, resulting in different executablesĮddy_openmp: This executable has been parallelised through OpenMP, which allows eddy to use more than one core/CPU when running.Įddy_cuda8.0: an eddy executable that has been parallelised with CUDA v8.0, which allows eddy to use an Nvidia GPU if one is available on the system.Įddy_cuda9.1: an eddy executable that has been parallelised with CUDA v9.1, which allows eddy to use an Nvidia GPU if one is available on the system.Įddy_cuda: On Linux, this is a convenience sym-link that points to the version of eddy_cuda that has been setup for you, or that you can configure for your system. Correcting susceptibility-by-movement interactions with eddyĮddy is a very computationally intense application, so in order to speed things up it has been parallelised.Correcting slice-to-volume movement with eddy.WARNING this page is being edited in preparation of a new release and may be in an inconsistent state.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |