 |
 |
Hi Darren, I have tried a few approaches, which haven't been successful on our 64 direction data. Some of the things that don't work: 1. taking the signal within the region of the artifact, and covarying for that signal in the fdt analysis (just got rid of most of the signal) 2. taking out the affected directions- for us, those seem to be any with x > .6 or < -.6, and left us with 42 directions. this got rid of the artifact, but decreased FA in the x direction fibers an average of 9% (tested in subjects without the artifact). In some people it ranged up to almost 25% signal loss. 3. we tried a version of the Gallichan correction but I'm not sure if it was quite right as it didn't fix the issue- it sounds the same as what you mentioned, subjects still had regions of artifact, although they seemed smaller. I'd also be very interested in hearing ways in which people have been able to successfully implement a fix for this! best, Katie On Mar 25, 2011, at 8:30 AM, Darren G wrote: > Dear FSL > > Has anyone implemented the Gallichan correction for the Siemens vibration artifact, and would you be willing to share your methods / scripts? We have tried the implementation and had some success, but our data is only 35 directions, and I think the correction is not as robust with fewer directions. Thus, we end up with some areas uncorrected; so we are particularly interested if you have applied the method to data with a similar number of directions. > > Thanks, > Darren _________________________________________ Katherine H. Karlsgodt, Ph.D. Center for Cognitive Neuroscience Department of Psychiatry University of California, Los Angeles email: [log in to unmask] phone: 310-206-3019 fax: 310-794-9740