"Error using spm_est_non_sphericity (line 196)
Please check your data: There are no significant voxels.

Hello Alfonso,

I hope you are doing well! We are trying to use the denoised data from CONN on SPM12 to run a simple first level analysis while taking advantage of aCompCor, and we are running into this error:
After reading about other similar issues, I ran the following potential fixes:

1) Changed the default level for FWE in the SPM defaults stats script from 0.001 to 0.05
2) Set the  high-pass filter on the 1st level specification in SPM to 'Inf'

Neither of these have worked in any combination. Just in case, this is what I'm doing:

1) Doing all preprocessing in CONN (default steps + ART). 
2) Running the Setup stage, with "Create confound effects Beta-maps" and "Create confound-corrected time-series" checked
- This is a simple block-design stories/noise task. 32 second blocks at TR = 2.
3) In the denoising step, I changed the band-pass filter to [0.008 inf], and removed the task effects from the confounds, leaving only ART outputs, CSF, and white matter.

After this, I go to the "/results/preprocessing/" folder and use all the volumes exported from the denoising step (the niftiDATA_Subject001_Condition000.nii file) for the first level analysis. For what it's worth, I also tried running the analysis without the denoising step on SPM12, and the activation is very strong, even with a stringent ART setup. 

Would you or anyone in the forum happen to know what I'm doing wrong? Any help would be very appreciated! Thank you!

Hi Claudio,

My guess is that this may be the effect of implicit masking. Since your data is already high-pass filtered the mean BOLD signal at each voxel is zero. SPM uses the mean BOLD signal at each voxel to determine "in-brain" vs. "out-of-brain" voxels so this may be causing some issues in your analyses. I would recommend to try setting the "masking threshold" value to "-inf" when defining your first-level analysis in SPM (its default value is "0.8") and see whether that helps.

Best
Alfonso
Originally posted by Claudio Toro:

Hi, Alfonso.

I met the same error as Claudio does. I have changed mask threshold to Inf but haven't fix it. Do you have other suggestions? Thanks!