questions > Using MRIcroGL and dcm2niix to convert CTs
Showing 1-5 of 5 posts
Dec 15, 2016 06:12 PM | TVASurg TVASurg
Using MRIcroGL and dcm2niix to convert CTs
Hi Dr. Rorden -
Many thanks for developing and releasing the MRIcroGL program as open-source.
We're trying to make use of the dcm2niix function to convert some of our hospital CT DICOMs to nii, however we're running into some interesting problems. The resulting nii files are quite numerous, and each file contains maybe 4-6 slices of the original set.
I've attached the dcm2niix log for your reference. What would you recommend in terms of troubleshooting?
Thanks again,
Alb
Many thanks for developing and releasing the MRIcroGL program as open-source.
We're trying to make use of the dcm2niix function to convert some of our hospital CT DICOMs to nii, however we're running into some interesting problems. The resulting nii files are quite numerous, and each file contains maybe 4-6 slices of the original set.
I've attached the dcm2niix log for your reference. What would you recommend in terms of troubleshooting?
Thanks again,
Alb
Dec 15, 2016 08:12 PM | Chris Rorden
RE: Using MRIcroGL and dcm2niix to convert CTs
Alb-
Short answer: Can you tell me if using the 'm'erge option works:
./dcm2niix -m y ~/DICOM/
Long answer: It looks like your vendor is incorrectly using DICOM tag 0018,1152 listing the slice number and not the exposure in milliamps. We use this tag to sort GE datasets (https://github.com/rordenlab/dcm2niix/pu...). Therefore, the ideal solution would be for your vendor to update their software to use this tag correctly. However, for a kludge, the 'm'erge command I mentioned above should override these values.
If you are able to send me an anonymized dataset, or acquire a dataset from a phantom image, I would be grateful if you could send me a sample (e.g. a dropbox link sent to my private email). There might be a way to auto-detect this error and deal with these images seamlessly.
Short answer: Can you tell me if using the 'm'erge option works:
./dcm2niix -m y ~/DICOM/
Long answer: It looks like your vendor is incorrectly using DICOM tag 0018,1152 listing the slice number and not the exposure in milliamps. We use this tag to sort GE datasets (https://github.com/rordenlab/dcm2niix/pu...). Therefore, the ideal solution would be for your vendor to update their software to use this tag correctly. However, for a kludge, the 'm'erge command I mentioned above should override these values.
If you are able to send me an anonymized dataset, or acquire a dataset from a phantom image, I would be grateful if you could send me a sample (e.g. a dropbox link sent to my private email). There might be a way to auto-detect this error and deal with these images seamlessly.
Dec 19, 2016 01:12 PM | TVASurg TVASurg
RE: Using MRIcroGL and dcm2niix to convert CTs
Dr. Rorden -
It works! Many many thanks. The merge option allows us to view our series in MRIcroGL64.
Looking for a way to share the dataset with you - as you mentioned our vendor doesn't seem to be using DICOM tags correctly, so I'm not sure if their anonymizing process is thorough enough. Would you recommend any processes/programs for anonymization?
Thanks,
Alb
It works! Many many thanks. The merge option allows us to view our series in MRIcroGL64.
Looking for a way to share the dataset with you - as you mentioned our vendor doesn't seem to be using DICOM tags correctly, so I'm not sure if their anonymizing process is thorough enough. Would you recommend any processes/programs for anonymization?
Thanks,
Alb
Dec 19, 2016 02:12 PM | Chris Rorden
RE: Using MRIcroGL and dcm2niix to convert CTs
DICOM is a very complex standard, and in my experience some
anonymizers can corrupt images. Two open source tools from very
experienced groups that are very popular (and therefore extensively
tested) are
1.) David Clunie's DicomCleaner (Windows OS only)
http://www.dclunie.com/pixelmed/software/webstart/DicomCleanerUsage.html
2.) Using the dcmtk's dcmodify function
https://gist.github.com/tinku99/2652845
1.) David Clunie's DicomCleaner (Windows OS only)
http://www.dclunie.com/pixelmed/software/webstart/DicomCleanerUsage.html
2.) Using the dcmtk's dcmodify function
https://gist.github.com/tinku99/2652845
Jan 23, 2017 03:01 PM | TVASurg TVASurg
RE: Using MRIcroGL and dcm2niix to convert CTs
Dr. Rorden -
Sorry for the late reply, I've been sick and have only recently gotten back to the office. We've prepared an anonymized DICOM series, and were wondering which email we should send it to?
Let us know, thanks again.
Alb
Sorry for the late reply, I've been sick and have only recently gotten back to the office. We've prepared an anonymized DICOM series, and were wondering which email we should send it to?
Let us know, thanks again.
Alb