Preprocessing

[Eric Kirby]

Hi Frank,

A few questions regarding preprocessing of some DWI data. We checked the quality of the raw dwi, eddy corrected dwi, motion corrected dwi, eddy & motion corrected dwi. I have attached an excel sheet of the quality control results.

1) Why is Eddy correction changing the image dimensions and increasing the number of bad slices?

2) Based on the attached excel, using only motion correction seems to be the best (high correlation and 0 bad slices), would you agree?... Then, how do we defend our choice of not correcting eddy currents in future publications and/or presentations (as eddy current correcting is a common practice)?

3) There is a preprocessing tick box under reconstruction. Should we just tick that instead of doing this motion and/or eddy correction? How do we check the quality of the results of doing that?

Looking forward to hearing from you!

Cheers,

Eric

[Frank Yeh]

Hi Eric,

    Sorry for my late reply:

A few questions regarding preprocessing of some DWI data. We checked the quality of the raw dwi, eddy corrected dwi, motion corrected dwi, eddy & motion corrected dwi. I have attached an excel sheet of the quality control results.

1) Why is Eddy correction changing the image dimensions and increasing the number of bad slices?

DSI Studio removed the background volume to speed up the process. The number of bad slices is only a reference and may not be a reliable estimation at the top or bottom slices.

Neighboring DWI correlation is a more reliable measurement.

 

2) Based on the attached excel, using only motion correction seems to be the best (high correlation and 0 bad slices), would you agree?... Then, how do we defend our choice of not correcting eddy currents in future publications and/or presentations (as eddy current correcting is a common practice)?

I would choose the eddy only.

The combination of eddy and additional motion correction only improves minimal (0.002).

 

3) There is a preprocessing tick box under reconstruction. Should we just tick that instead of doing this motion and/or eddy correction? How do we check the quality of the results of doing that?

No, I would suggest users choose the best approach for their data because not all data work with the default preprocessing pipeline.

Best regards,

Frank

 

Looking forward to hearing from you!

Cheers,

Eric

--
You received this message because you are subscribed to the Google Groups "DSI Studio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dsi-studio+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/116a65e7-9a1e-427a-872a-852a54c1904en%40googlegroups.com.