"Applying" results of DRBUDDI to full dataset

Submitted by randomwalkerdi… on


New user to TORTOISE!

I have diffusion data with 7 up and 7 direction, and a full data set with 64 directions. I have been able to follow the steps till DRBUDDI stage; I have provided the 7 direction blip up data and the 7 direction blip down data. Once I have the results, how can I "apply" the results of distortion correction to my full data? Or do I need to run DRBUDDI with full data (blip up) and 7 directions blip down?

Apologies for the newbie questions.



Thank you!

Keep diffusing!

Submitted by nayaka on Fri, 04/19/2019 - 12:14



Welcome to TORTOISE land.

Just to clarify your question, do you mean you have up data with 64 volumes and down data with 7 volumes? If 7 volume down data- are these DWI's+b0 or just b0's?

The following answer applies to the scenario where you have unequal volumes between up and down data:

Please provide the up data with 64 volumes (i.e up_proc.list) and down data with 7 volumes (i.e down_proc.list) to DRBUDDI. It will complain at the beginning with a warning message indicating that there is a mismatch in volumes but it will continue with the processing. In your DRBUDDI_proc folder, the final corrected result will not be one final DRBUDDI.list, nii and bmtxt file as typically expected with up down matched volume dataset. In this case you will have DRBUDDI.list, .nii and .bmtxt, each, for both up and down data. The corrections will be applied to the up data using information from the down data. So, please use the up_DRBUDDI corrected list file from the DRBUDDI_proc folder in your tensor fitting.



Submitted by randomwalkerdi… on Sun, 04/21/2019 - 06:54



Thank you for replying and the welcome!

I have one up data consisting of one b=0 and 6 b=1000 volumes. I have down data consisting of one b=0 and 6 b=1000 volumes. Finally, I have my "full" data with 1 b=0 and 64 b=1000 volumes (this is up direction).

Previously, I used the up and down data (1 b=0 and 6 b=1000) and passed it on to DRBUDDI resulting in corrected data. However, I wanted to know if there is a way to estimate the correction from these smaller dataset and then apply it to the full data set? Since the acquisition details were exactly same for all three sets of acquisition (including factors like EPI readout time), I was hoping that one could calculate the distortion from the smaller data set and then apply it to the full one.

If that is not possible, then I guess I can use the 7 volume down data and the full data in the way that you mentioned.


Thanks and keep diffusing!