Back to the main page.
Bug 1367 - slow speed of niftii file creation using sourceinterpolate.m for 4D datasets
| Status | CLOSED WONTFIX |
| Reported | 2012-03-12 14:49:00 +0100 |
| Modified | 2014-03-06 15:39:29 +0100 |
| Product: | FieldTrip |
| Component: | core |
| Version: | unspecified |
| Hardware: | Macintosh |
| Operating System: | Mac OS |
| Importance: | P3 enhancement |
| Assigned to: | Jan-Mathijs Schoffelen |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
woolrich - 2012-03-12 14:49:48 +0100
When working with 4D datasets, e.g. 3 dimensions of space a a grid resolution of 5mm and 200 time points, the call to sourceinterpolate.m to output the niftii volumes for all 200 volumes is painfully slow. I have implemented my own approach to setting up and handling beamformer grids, and for outputting niftii files to overcome this problem. However this makes use of tools in FSL, and it would be great for the Fieldtrip community if a solution to this was available within Fieldtrip itself. Many thanks, Mark Woolrich.
Jan-Mathijs Schoffelen - 2014-01-15 14:56:07 +0100
Reviving some oldies: Hi Mark, Is this still an issue? What was exactly the case from the top of your head ;-)? Was the interpolation slow, or the writing to disk? I think nowadays we support writing of 4D volumes to nifti using ft_volumewrite, which should work pretty straightforwardly, and there wouldn't be a need to call ft_sourceinterpolate to begin with (unless you'd want to upsample your grids).