mri_convert ________________________________________ Von: freesurfer-bounces@nmr.mgh.harvard.edu freesurfer-bounces@nmr.mgh.harvard.edu im Auftrag von dgw dgwakeman@gmail.com Gesendet: Donnerstag, 28. Januar 2016 15:55 An: Freesurfer support list Betreff: Re: [Freesurfer] Segmentation depends on image file format

What software did you use to generate the NIFTI files? Many of the converters do different things to the axes of the data, and this does affect results.

hth d

On Thu, Jan 28, 2016 at 8:49 AM, Parzer, Peter Peter.Parzer@med.uni-heidelberg.de wrote:

Hi,

we have brain images stored as DICOM and as NIFTI files. For some time we used to import the NIFTI files to freesurfer for segmentation. I was curious if it makes a difference if we would import the DICOM files directly, and was surprised that it actually did. Here are the commands we used:

# Import DICOM files recon-all -s 3003PC_dcm -i ../../Daten/DICOM/3003PC/M_MPRAGE_sag/IM-0010-0001.dcm

# Convert DICOM to NIFTI and import NIFTI file mri_convert ../../Daten/DICOM/3003PC/M_MPRAGE_sag/IM-0010-0001.dcm 3003PC.nii.gz recon-all -s 3003PC_nii -i 3003PC.nii.gz

# Segmentation of DICOM file import recon-all -s 3003PC_dcm -all

# Segmentation of NIFTI file import recon-all -s 3003PC_nii -all

Here are the first 3 volumes from aseg.stats for 3003PC_dcm (DICOM import)

1 4 5612 5611.8 Left-Lateral-Ventricle 27.6355 12.8465 8.0000 93.0000 85.0000 2 5 452 452.2 Left-Inf-Lat-Vent 42.1100 14.9377 7.0000 86.0000 79.0000 3 7 13765 13765.1 Left-Cerebellum-White-Matter 87.7808 6.6008 36.0000 109.0000 73.0000

and here the the first 3 volumes for 3003PC_nii (NIFTI import)

1 4 5582 5582.2 Left-Lateral-Ventricle 27.4507 12.7212 8.0000 93.0000 85.0000 2 5 421 421.1 Left-Inf-Lat-Vent 40.8032 14.6554 7.0000 83.0000 76.0000 3 7 13899 13898.6 Left-Cerebellum-White-Matter 87.6707 6.7539 36.0000 109.0000 73.0000

The differences are small but I think there should be no differences at all. Has someone an explanation for this differences and which file format is better?

Greetings Peter

P. S.: Our system: freesurfer-Linux-centos6_x86_64-stable-pub-v5.3.0 on Ubuntu 14.04 _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.

_______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

I suspect that it has to do with the image geometry information. In nifit, this is stored in quaterion format. In FS, this is converted to direction cosines. This conversion is not reversible. So when you mri_convert the DICOMS to nii, mri_convert will get the direction cosines from the dicom file and convert them to quaternions. When you run recon-all using the nii file, the quaternions are converted back to direction cosines. But this does not yield exactly the same as what was in the dicom file. This conversion does not happen when you start out with mgz files. In version 6 I changed the code to try to reduce this effect, but I think it will always be there.

On 01/28/2016 10:49 AM, Bruce Fischl wrote:

can you run mri_diff on the orig.mgz in the two runs and send us the output? On Thu, 28 Jan 2016, Parzer, Peter wrote:

...

mri_convert ________________________________________ Von: freesurfer-bounces@nmr.mgh.harvard.edu freesurfer-bounces@nmr.mgh.harvard.edu im Auftrag von dgw dgwakeman@gmail.com Gesendet: Donnerstag, 28. Januar 2016 15:55 An: Freesurfer support list Betreff: Re: [Freesurfer] Segmentation depends on image file format

What software did you use to generate the NIFTI files? Many of the converters do different things to the axes of the data, and this does affect results.

hth d

On Thu, Jan 28, 2016 at 8:49 AM, Parzer, Peter Peter.Parzer@med.uni-heidelberg.de wrote:

...

Hi,

we have brain images stored as DICOM and as NIFTI files. For some time we used to import the NIFTI files to freesurfer for segmentation. I was curious if it makes a difference if we would import the DICOM files directly, and was surprised that it actually did. Here are the commands we used:

# Import DICOM files recon-all -s 3003PC_dcm -i ../../Daten/DICOM/3003PC/M_MPRAGE_sag/IM-0010-0001.dcm

# Convert DICOM to NIFTI and import NIFTI file mri_convert ../../Daten/DICOM/3003PC/M_MPRAGE_sag/IM-0010-0001.dcm 3003PC.nii.gz recon-all -s 3003PC_nii -i 3003PC.nii.gz

# Segmentation of DICOM file import recon-all -s 3003PC_dcm -all

# Segmentation of NIFTI file import recon-all -s 3003PC_nii -all

Here are the first 3 volumes from aseg.stats for 3003PC_dcm (DICOM import)

1 4 5612 5611.8 Left-Lateral-Ventricle 27.6355 12.8465 8.0000 93.0000 85.0000 2 5 452 452.2 Left-Inf-Lat-Vent 42.1100 14.9377 7.0000 86.0000 79.0000 3 7 13765 13765.1 Left-Cerebellum-White-Matter 87.7808 6.6008 36.0000 109.0000 73.0000

and here the the first 3 volumes for 3003PC_nii (NIFTI import)

1 4 5582 5582.2 Left-Lateral-Ventricle 27.4507 12.7212 8.0000 93.0000 85.0000 2 5 421 421.1 Left-Inf-Lat-Vent 40.8032 14.6554 7.0000 83.0000 76.0000 3 7 13899 13898.6 Left-Cerebellum-White-Matter 87.6707 6.7539 36.0000 109.0000 73.0000

The differences are small but I think there should be no differences at all. Has someone an explanation for this differences and which file format is better?

Greetings Peter

P. S.: Our system: freesurfer-Linux-centos6_x86_64-stable-pub-v5.3.0 on Ubuntu 14.04 _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer