Slice Number and Slice order

Submitted by florence.ruby on

Hello, 

I used DPARSF Basic a few weeks ago to preprocess data from the Enhanced NKI Rockland Sample.

Recently I checked the Autosave .mat file and found that the Slice Number I had indicated was wrong, I had put 33 instead of 38. 
However, the slice order ([1 3 5... 34 36 38]) and slice reference (38) were approriate and the prepocessing ran smoothly. 
I will run the preprocessing again with the correct parameter but I still would like to know in what way can the wrong slice number affect the preprocessing steps? How can I compare results from the preprocessing with the wrong and correct slice number? 

Thank you for your help! 

Florence

 

 

 

YAN Chao-Gan

Fri, 01/10/2014 - 20:04

Hi,

If you put slice number as 33 instead of 38, you should got an error.

I will suggest you re-run it, or skip slice timing if you don't know the correct slice order -- although slice timing seems should not have a very severe effect. You can compare the resists with paired t-test on the derivatives (e.g., functional connectivity maps) with correct order and those with incorrect order.

Best,

Chao-Gan

 

florence.ruby

Mon, 01/20/2014 - 21:55

In reply to by YAN Chao-Gan

Hi!

Thanks for your answer! 

I tested the preprocessing with different parameters and reran it on a single participant's data to see what would make DPARSF produce an error. Only when the slice order is wrong (i.e. not up to 38 in m case) will there be an error. The slice number does not have any effect apparently (preprocessing runs whatever number I put in). 
I compared the results of functional connectivity for that subject for different slice numbers and they are all identical.
So this made me think the only parameter DPARSF is actually using is the slice order and not slice number. 
But in any case I will rerun all the participants with correct parameters. 

Thanks again!

Florence

 

YAN Chao-Gan

Fri, 01/24/2014 - 20:19

In reply to by florence.ruby

Hi Florence,

Thanks a lot for your follow-up.

Happy to know that the slice number doesn't make any difference. DPARSF just transfer the settings to SPM for slice timing, seems SPM only depends on the slice order information other than slice number (since the number is implicit within the order).

Best,

Chao-Gan