search

TheAxonLab / hcph-sops

SOPs of the HCPh project
https://www.axonlab.org/hcph-sops/
Creative Commons Attribution 4.0 International
0 stars 6 forks source link

Unidentified artifact #366

Open celprov opened 12 months ago

celprov commented 12 months ago

Session 15 of the piloting phase presents a weird square in the sagittal view of the standard deviation map.

I'm unsure what to do with this information, but I'm putting it here so we can remember and maybe spark discussion about what this is.

image

celprov commented 11 months ago

@oesteban I have a bad news ^^' ^^' All the fMRI scans are suffering from this artifact, but with different intensity, sometimes it's very clear as above (e.g ses-25) , sometimes it is quite subtle (e.g ses-23). The intensity of the artifact is consistent within one session (meaning that qct, bht and rs within one session suffer from similar relative artifact intensity) Maybe this artifact is specific to echo1? Can you run another echo so we can assess whether the artifact is present there too?

celprov commented 11 months ago

@oesteban suggested a sliding-window temporal SNR to check whether this high variability comes from frames close to the end or beginning of the scan.

acionca commented 11 months ago

I may have a look if it mostly happens at similar time points.

acionca commented 11 months ago

Alright, I computed the standard deviation in the temporal axis (I believe it is what is shown in your MRIQC report) at different time windows (10 windows of size 75 timepoints).

Bad news, the rectangle appears similarly in all windows (see below for ses_001 echo_1 and echo_2)...

sub-001_ses-001_task-rest_dir-LR_echo-12_part-mag_desc-windowedSTD_bold-MOD

I have never seen this before but my first guess (and I might be wrong) is that it looks like two images covering the lower and upper 2/3rd of the brain respectively are overlayed (see below). Could it be possible ? Maybe an issue during the reconstruction ?

one_full_std-rectangle