Question
Question Posted 01/09/20:
We are interested in investigating how TAU signal varies over the course of the 30 minute acquisition. To do so, I would like some assurance that the alphanumeric order that is produced when you (ADNI) convert them into NiFTI (6ea. filenames) is also the frame order. I was expecting a multiframe image, but you appear to parse them into single 5min. frames, and I'm worried that the long string of information in the middle of a filename might jumble the frame order.
ADNI_xxx_S_xxxx_PT_ADNI_Brain_PET__Raw_br_raw_20160216135803605_530_SERIESUID_IMAGEUID.nii
ADNI_xxx_S_xxxx_PT_ADNI_Brain_PET__Raw_br_raw_20160216135804924_23_SERIESUID_IMAGEUID.nii
ADNI_xxx_S_xxxx_PT_ADNI_Brain_PET__Raw_br_raw_20160216135806224_138_SERIESUID_IMAGEUID.nii
ADNI_xxx_S_xxxx_PT_ADNI_Brain_PET__Raw_br_raw_20160216135807524_416_SERIESUID_IMAGEUID.nii
ADNI_xxx_S_xxxx_PT_ADNI_Brain_PET__Raw_br_raw_20160216135814650_357_SERIESUID_IMAGEUID.nii
ADNI_xxx_S_xxxx_PT_ADNI_Brain_PET__Raw_br_raw_20160216135817897_201_SERIESUID_IMAGEUID.nii
For example
ADNI_xxx_S_xxxx_PT_ADNI_Brain_PET__Raw_br_raw_ "20160216135" (internally consistent)..."803605_530" (varies and dictates the alpha order) _SERIESUID_IMAGEUID.nii
Also, for future reference, I am assuming that this subject's scan was processed on 2/16/2016 (20160216), and that the rest for the string is a more granular time-stamp. That is, 135803605 refers to 13:58 (1:58PM in the afternoon) with the final five characters being some kind of explicit "second/miliseconds" report. Am I correct with that assumption?
While I assume if I'd downloaded these as DICOM, frame order would be part of the header and therefore maintained when/if I converted to a multiframe NiFTI. However, I'd prefer to simply use your NiFTI's.
We are interested in investigating how TAU signal varies over the course of the 30 minute acquisition. To do so, I would like some assurance that the alphanumeric order that is produced when you (ADNI) convert them into NiFTI (6ea. filenames) is also the frame order. I was expecting a multiframe image, but you appear to parse them into single 5min. frames, and I'm worried that the long string of information in the middle of a filename might jumble the frame order.
ADNI_xxx_S_xxxx_PT_ADNI_Brain_PET__Raw_br_raw_20160216135803605_530_SERIESUID_IMAGEUID.nii
ADNI_xxx_S_xxxx_PT_ADNI_Brain_PET__Raw_br_raw_20160216135804924_23_SERIESUID_IMAGEUID.nii
ADNI_xxx_S_xxxx_PT_ADNI_Brain_PET__Raw_br_raw_20160216135806224_138_SERIESUID_IMAGEUID.nii
ADNI_xxx_S_xxxx_PT_ADNI_Brain_PET__Raw_br_raw_20160216135807524_416_SERIESUID_IMAGEUID.nii
ADNI_xxx_S_xxxx_PT_ADNI_Brain_PET__Raw_br_raw_20160216135814650_357_SERIESUID_IMAGEUID.nii
ADNI_xxx_S_xxxx_PT_ADNI_Brain_PET__Raw_br_raw_20160216135817897_201_SERIESUID_IMAGEUID.nii
For example
ADNI_xxx_S_xxxx_PT_ADNI_Brain_PET__Raw_br_raw_ "20160216135" (internally consistent)..."803605_530" (varies and dictates the alpha order) _SERIESUID_IMAGEUID.nii
Also, for future reference, I am assuming that this subject's scan was processed on 2/16/2016 (20160216), and that the rest for the string is a more granular time-stamp. That is, 135803605 refers to 13:58 (1:58PM in the afternoon) with the final five characters being some kind of explicit "second/miliseconds" report. Am I correct with that assumption?
While I assume if I'd downloaded these as DICOM, frame order would be part of the header and therefore maintained when/if I converted to a multiframe NiFTI. However, I'd prefer to simply use your NiFTI's.
Response posted 01/09/20 by KC:
We recommend you download the image in their native format (DICOM or ECAT) and translate to NIfTI using your favorite program.
Response posted 01/09/20 by Randolph Andrews:
So there's no guarantee that the NiFTIs you are providing via the LONI interface will not not be jumbled in frame order if sorted alphanumerically?
Response posted 01/09/20 by Bob Koeppe:
The translation from DICOM to Nifti is done at LONI, so any problems with frame order is happening at this step. I second the option to download the images as DICOM and do the conversion to nifti using your own routines. There is not problem with the DICOM provided by LONI
