help > Classification of DICOM images
Showing 1-2 of 2 posts
Feb 10, 2022 04:02 AM | Krystal Yau
Classification of DICOM images
Dear expert,
I have been searching up online on how to classify the DICOM images into different categories/ folders, i.e. T1, T2, FLAIR, M0 and PCASL.
Currently, the files that named with Z in the beginning, e.g. "Z01", "Z02" ... "Z8568". After converting dcm images into nii files are basically named "'myFolder_MPRAGE_19770703150928_1.nii'". I understand that I can modify the file name by modifying the annotation under "Output Filename". I'm stuck at the step of categorizing them...
Could you please advise how to classify the images as well as in batch processing?
Your attention and help will be highly appreciated!
Thanks a lot.
Best regards,
Krystal
I have been searching up online on how to classify the DICOM images into different categories/ folders, i.e. T1, T2, FLAIR, M0 and PCASL.
Currently, the files that named with Z in the beginning, e.g. "Z01", "Z02" ... "Z8568". After converting dcm images into nii files are basically named "'myFolder_MPRAGE_19770703150928_1.nii'". I understand that I can modify the file name by modifying the annotation under "Output Filename". I'm stuck at the step of categorizing them...
Could you please advise how to classify the images as well as in batch processing?
Your attention and help will be highly appreciated!
Thanks a lot.
Best regards,
Krystal
Feb 10, 2022 12:02 PM | Chris Rorden
RE: Classification of DICOM images
dcm2niix does not know the intention of your sequences. A T2* scan
might be a task fMRI in one series and a resting state acquisition
in another.
You may want to look at one of the many tools that wraps dcm2niix with your study-specific heuristics. Examples include ezBIDS, Heidiconv, and Dcm2Bids
https://github.com/rordenlab/dcm2niix#li...
For future acquisitions, you can set the series name on your console to match your intention, e.g. "T1", "T2", "FLAIR". You can even do this to help automate conversion to BIDS data:
https://github.com/ReproNim/reproin
You may want to look at one of the many tools that wraps dcm2niix with your study-specific heuristics. Examples include ezBIDS, Heidiconv, and Dcm2Bids
https://github.com/rordenlab/dcm2niix#li...
For future acquisitions, you can set the series name on your console to match your intention, e.g. "T1", "T2", "FLAIR". You can even do this to help automate conversion to BIDS data:
https://github.com/ReproNim/reproin
