I would like to define ROIs of WM tracts in a neonatal image of brain. I have 3dT1 and DKI scan. I should use 3dT1 as an anatomical reference. Do you suggest any existing neonatal atlas to do so?
thank you.
Here you find attached my data
https://onedrive.live.com/?id=B0C0561E34E9D76F!613&cid=B0C0561E34E9D76F
Hi,
There are two neonatal parcellations of WM that I’m aware:
This one contains: T2w,T1w and FA and can be found here:
http://cmrm.med.jhmi.edu/cmrm/Data_neonate_atlas/atlas_neonate.htm
And the second one is based on the M-CRIB atlas, but I don’t know if it is available:
The rest of the atlas I’m aware of, don’t have WM parcellation to use without T2w. I hope this helps.
Best regards,
Manuel
2 Likes
Thank you. Do you just know how to open .dat file ?
thank you
Rosella
Hi,
I don’t know that format, I think is an extension to be used in their own software. In the past I used with the analyze version.
Best regards,
Manuel
1 Like
I think this might help:
https://www.nitrc.org/forum/forum.php?thread_id=2118&forum_id=2
yes,
thank you I have just seen that post.
Rosella
Hi @ros,
I am curious about this, is there any difference if you convert the .dat to nifti or the analyze to nifti? Are the same images?
Best regards,
Manuel
I think they are the same images! However, some images are there just in .dat format as far as I can see.
Rosella
Hi. Do you know if JHU neonate atlas is integrated with FSL?
thanks
Rosella
Hi,
Do you know if JHU neonate atlas is integrated with FSL?
I don’t think so, I think they have the adult atlas only.
Best regards,
Manuel
Ok because I would need to calculate some DTI metrics over each of the 122 ROis contained in the atlas parcellation , but I have to find a quick way to extract each ROI separately and use it as a mask .
thanks,
Rosella
Hi,
You could do that from the command line, for example for the label 2:
fslmaths ID_LABELS_B0space.nii.gz -thr 2 -uthr 2 -bin ID_label_2.nii.gz
fslstats ID_FA.nii.gz -k ID_label_2.nii.gz -m
or with mrtrix:
mrstats -output mean -mask ID_label_2.nii.gz ID_FA.nii.gz
Then you just need to do it in a for loop (this is how I do it). I hope this helps.
Best regards,
Manuel
Yes, for your atlas this is ok and I use it too but for the JHU atlas I do not have separate image corresponding to each ROI.
Thanks
Rosella
HI,
The firs command I wrote, generates an independent image for each ROI (you just need to replace the 2 for the ID of the other ROIs).
fslmaths ID_LABELS_B0space.nii.gz -thr 2 -uthr 2 -bin ID_label_2.nii.gz
Best regards,
Manuel
Ah ok thank you! And ID_label_B0_space.nii.gz is the parcellation map image, isn’t it?
many thanks
Rosella
Hi,
Yes, registered to your diffusion image.
Best regards,
Manuel
1 Like
Thank you very much!
Rosella
Hi,
I am currently working on neonatal image registration using the JHU_neonate_SS atlas. Could you please share it with me? I am unable to access the download website: http://cmrm.med.jhmi.edu/cmrm/Data_neonate_atlas/atlas_neonate.htm
Thank you!
Best regards,
Hanjing
Hi,
You are right, the link doesn’t seem to work. However, the authors have a new tool that they recommend to use (still beta version and being tested):
The main method for adult brain is published here
If you still want the atlas, I should have it in my computer from previous experiments, please drop me an email (manuel.blesa@ed.ac.uk) and I’ll send it to you.
Best regards,
Manuel