questions
58 Subscribers 
questions > the difference of created 3 files
Showing 1-2 of 2 posts
Jan 2, 2018 01:01 AM | Kyungil Park
the difference of created 3 files
Hi, all
Please understand my naive question.
When I ran dicom2nii with 150 dicom files, 3 different files of nii format were created.
I am wondering how different are these 3 files?
Thank you so much in advance for your help.
Kyung
Please understand my naive question.
When I ran dicom2nii with 150 dicom files, 3 different files of nii format were created.
I am wondering how different are these 3 files?
Thank you so much in advance for your help.
Kyung
Jan 2, 2018 06:01 PM | Chris Rorden
RE: the difference of created 3 files
Hello-
First of all, if you are a new user, I would suggest using my modern dcm2niix (which comes with MRIcroGL) over my older dcm2nii (which comes with MRIcron). Vendors update there formats very quickly, so using the latest tool is a good idea.
Second, the results you get depend on your input data. Consider that your 150 dicom files include a 3D T1-weighted image with 100 slices (saved as 2D DICOM images) and two 4D fMRI runs each with 25 volumes (each saved as 3D DICOM mosaics). In this case, your 3 volumes would be a T1 scan and two fMRI scans. You can usually visually inspect your images to determine the modality. However, I usually convert my DICOM images with a command like this "dcm2niix -f %p_%t_%s ~/myDICOMfolder". In this case, the resulting files will have the protocol name (%p), the study time (%t) and series number (%s) added. If you acquired your data with sensible protocol names, then your images should be named something like "T1_201801011023_2.nii", "fMRI_201801011023_2.nii", "fMRI_201801011023_2.nii", so it is pretty easy to identify the modality. Note dcm2niix can also generate BIDS-format text files if you provide the "-b y" argument: "dcm2niix =b y -f %p_%t_%s ~/myDICOMfolder". For example, you would get a file named "T1_201801011023_2.json" that will provide a lot of details about the sequence.
I hope this helps. The DICOM standard is very complex, so one needs to know the type of images you acquired to provide a good answer. However, please click on the "MediaWiki" button on the left for the manual.
First of all, if you are a new user, I would suggest using my modern dcm2niix (which comes with MRIcroGL) over my older dcm2nii (which comes with MRIcron). Vendors update there formats very quickly, so using the latest tool is a good idea.
Second, the results you get depend on your input data. Consider that your 150 dicom files include a 3D T1-weighted image with 100 slices (saved as 2D DICOM images) and two 4D fMRI runs each with 25 volumes (each saved as 3D DICOM mosaics). In this case, your 3 volumes would be a T1 scan and two fMRI scans. You can usually visually inspect your images to determine the modality. However, I usually convert my DICOM images with a command like this "dcm2niix -f %p_%t_%s ~/myDICOMfolder". In this case, the resulting files will have the protocol name (%p), the study time (%t) and series number (%s) added. If you acquired your data with sensible protocol names, then your images should be named something like "T1_201801011023_2.nii", "fMRI_201801011023_2.nii", "fMRI_201801011023_2.nii", so it is pretty easy to identify the modality. Note dcm2niix can also generate BIDS-format text files if you provide the "-b y" argument: "dcm2niix =b y -f %p_%t_%s ~/myDICOMfolder". For example, you would get a file named "T1_201801011023_2.json" that will provide a lot of details about the sequence.
I hope this helps. The DICOM standard is very complex, so one needs to know the type of images you acquired to provide a good answer. However, please click on the "MediaWiki" button on the left for the manual.