[RESOLVED] T1w QC shows data quality not sufficient to proceed

Use Questions
1

Description:
When preprocessing my dataset after running T1w QC the data quality does not look good enough to proceed. Is there anything I can do at this stage to try and improve the data quality?

Call:
Preprocessing script so far:

# -- PreFreesurfer hcp_pre_freesurfer
qunex hcp_pre_freesurfer \
	--sessions="/gpfs/project/fas/n3/Studies/Psi.Med.2/processing/batch_psi_med.txt" \
	--sessionsfolder="/gpfs/project/fas/n3/Studies/Psi.Med.2/sessions/" \
	--overwrite=no \
	--parsessions=10 \
	--hcp_t2=NONE \
 	--scheduler="SLURM,time=2-00:00:00,ntasks=4,cpus-per-task=2,mem-per-cpu=100000,partition=pi_anticevic_bigmem"


# -- Freesurfer hcp_freesurfer 
qunex hcp_freesurfer \
	--sessions="/gpfs/project/fas/n3/Studies/Psi.Med.2/processing/batch_psi_med.txt" \
	--sessionsfolder="/gpfs/project/fas/n3/Studies/Psi.Med.2/sessions/" \
	--overwrite=no \
	--parsessions=1 \
	--hcp_t2=NONE \
 	--scheduler="SLURM,time=2-00:00:00,ntasks=1,cpus-per-task=2,mem-per-cpu=100000,partition=pi_anticevic_bigmem"

# -- PostFreesurfer hcp_post_freesurfer
qunex hcp_post_freesurfer \
	--sessions="/gpfs/project/fas/n3/Studies/Psi.Med.2/processing/batch_psi_med.txt" \
	--sessionsfolder="/gpfs/project/fas/n3/Studies/Psi.Med.2/sessions/" \
	--overwrite=no \
	--parsessions=1 \
	--hcp_t2=NONE \
 	--scheduler="SLURM,time=2-00:00:00,ntasks=1,cpus-per-task=2,mem-per-cpu=100000,partition=pi_anticevic_bigmem"

# -- T1w QC
qunex run_qc \
	--sessionsfolder="/gpfs/project/fas/n3/Studies/Psi.Med.2/sessions/" \
	--sessions="P124_5107_M1,P124_5107_M2,P124_5115_M1,P124_5115_M2,P124_5116_M1,P124_5116_M2,P124_5118_M1,P124_5123_M1,P124_5123_M2,P124_5127_M1,P124_5127_M2,P124_5128_M1,P124_5128_M2,P124_5130_M1,P124_5130_M2,P124_5131_M1,P124_5131_M2,P124_5132_M1,P124_5132_M2,P124_5133_M1,P124_5133_M2,P124_5136_M1,P124_5136_M2,P124_5141_M1,P124_5141_M2,P124_5142_M1,P124_5142_M2,P124_5144_M1,P124_5144_M2,P124_5146_M1,P124_5146_M2,P124_5147_M1,P124_5147_M2,P124_5151_M1,P124_5151_M2,P124_5152_M1,P124_5152_M2,P124_5153_M1,P124_5153_M2,P124_5155_M1,P124_5155_M2,P124_5156_M1,P124_5156_M2,P124_5157_M1,P124_5157_M2,P124_5158_M1,P124_5158_M2,P124_5159_M1,P124_5159_M2,P124_5160_M1,P124_5170_M2" \
	--outpath='/gpfs/project/fas/n3/Studies/Psi.Med.2/sessions/QC' \
	--scenetemplatefolder='/gpfs/project/fas/n3/software/qunex/qx_library/data/scenes/qc' \
	--modality='T1w' \
	--overwrite='yes' \
 	--scheduler="SLURM,time=2-00:00:00,ntasks=1,cpus-per-task=2,mem-per-cpu=100000,partition=pi_anticevic"

Logs:
T1w QC images are in: /gpfs/project/fas/n3/Studies/Psi.Med.2/sessions/QC
All logs are in: /gpfs/project/fas/n3/Studies/Psi.Med.2/processing/logs/comlogs

2

Hi Flora,
I don’t know if there is much you can do to change the quality of the data at this point, but it may be possible to improve the FreeSurfer segmentation/surface tracing results.

We are also currently working on supporting two other more manually intensive methods:

If you need these options, I will need to check with the rest of the team about when these will be ready and can get back to you in the next few days.

3

Hi Flora, I conferred with the dev team and the other edits should be able to be passed to FreeSurfer via the --hcp_fs_extra_reconall parameter in qx hcp_freesurfer.

e.g. for control points, you can save the control points file in:
/gpfs/project/fas/n3/Studies/Psi.Med.2/sessions/<sessionID>/hcp/<sessionID>/T1w/<sessionID>/tmp/control.dat

Then, you can pass --hcp_fs_extra_reconall="-autorecon2-cp|-autorecon3" (as per FsTutorial/ControlPoints_freeview - Free Surfer Wiki).