Hi Neuro-starts,
In attempts to bidsdify a public dataset (e.g. PPMI), I realize that the SeriesDescription field could be different for specific sequences (in this example, T1w), which I would like to BIDS-organize it like …/*/anat/*T1w.nii.gz :
(dcm2bids) $ for DIR in $(ls -1d ./*/1/*/ ) ; do AA=$(find ${DIR}/ -maxdepth 1 -name "*.dcm" | tail -1) ; echo $AA ; dcmdump $AA | grep SeriesDescription ; done
.
.
.
./101026/1/I1498714//PPMI_101026_MR_3D-T1-weighted_SAGITAL_br_raw_20210930191033267_124_S1068250_I1498714.dcm
(0008,103e) LO [3D-T1-weighted_SAGITAL]
.
.
.
./101179/1/I1493072//PPMI_101179_MR_3D_T1-weighted_br_raw_20210916223305322_1_S1063545_I1493072.dcm
(0008,103e) LO [3D T1-weighted]
I am having issues specifying a way to correctly define multiple SeriesDescription patterns in the conf. file. Is this even possible? If so, I would appreciate any guidance.
This example will work only for the first pattern:
{
"descriptions":[
{
"dataType": "anat",
"modalityLabel": "T1w",
"criteria": {
"SeriesDescription": "3D-T1-weighted_SAGITAL"
}
},
}
However, when I tried to add multiple SeriesDescription patterns as a list:
{
"descriptions":[
{
"dataType": "anat",
"modalityLabel": "T1w",
"criteria": {
"SeriesDescription": ["3D-T1-weighted_SAGITAL", "3D T1-weighted"]
}
},
}
Or as a nested elements [most likely wrong]:
{
"descriptions":[
{
"dataType": "anat",
"modalityLabel": "T1w",
"criteria": {
"SeriesDescription": "3D-T1-weighted_SAGITAL",
"SeriesDescription: "3D T1-weighted"
}
},
}
I don’t get the pattern recognized. I considered looking at other parameters (e.g. InversionTime, EchoTime, …) yet there is more variability within.
Any helpful suggestions will be appreciated. I also found https://github.com/rmarkello/ppmitobids yet the information there is deprecated and not working accordingly.
Regards,
Rodrigo