Closed PaulBautin closed 4 years ago
@PaulBautin you need to click inside the spinal cord:
@PaulBautin i can see that some data are of very poor quality-- can you please start modality-specific issues about data quality as @valosekj did in #14?
from what i see we can exclude:
@PaulBautin i'm looking at both zip files, and some images appear in both zip files, which:
could you please re-upload a proper qc, without duplication? thanks
I have corrected labeling, see QC: qc_corr_20200730115633.zip
Is there a reason not to allow "none type" input in a section of the config yaml file? sg_manual_correction did not allow to only do labelling corrections.
I have corrected labeling, see QC: qc_corr_20200730115633.zip
the qc also contains segmentations, which is confusing
Is there a reason not to allow "none type" input in a section of the config yaml file? sg_manual_correction did not allow to only do labelling corrections.
just remove the FILES_SEG and FILES_GMSEG entries in your config file-- or you can also open an issue so we can fix it at some point
Is there a reason not to allow "none type" input in a section of the config yaml file? sg_manual_correction did not allow to only do labelling corrections.
just remove the FILES_SEG and FILES_GMSEG entries in your config file-- or you can also open an issue so we can fix it at some point
It is probably same thing as I mentioned in this issue. I would vote for clarifying doc -- add a sentence/hint something like: "If you want to correct only labeling, include in your config yml file only FILES_LABEL
key, i.e., do not include/delete FILES_SEG
and FILES_GMSEG
keys, and vice versa."
new QC with command: sg_manual_correction -config err_file_t1w_t2w_corrected.yml -path-in data_to_correct -qc-only
qc_corr_20200731105752.zip
@PaulBautin can you pls remove sub-cmrra02_T2w_RPI_r.nii.gz
from the manual correction, because there is the problem with the data, i.e., this should not be part of the "manual correction" PR.
@PaulBautin also remove sub-sapienza04_T2star_rms.nii.gz
, where the segmentation is meaningless because of poor data quality.
also: this manual corr is not complete:
DONE:
new QC: qc_corr_20200731122832.zip
This PR contains manually corrected t1w and t2w SC segmentations for following subjects:
FILES_SEG:
FILES_GMSEG:
FILES_LABEL:
-Very little correction was applied to GMSEG because of the difficulty to localize grey matter -sub-cmrra02_T2w_RPI_r.nii.gz was not corrected because of defacing problem
Below is zipped QC report: (Second zip file contains corrections of first file for): FILES_SEG:
FILES_GMSEG:
FILES_LABEL:
qc_corr_20200729145633.zip qc_corr_20200730095951.zip