Template space for recobundles: ICMB 2009a vs. ICBM 2009c

yeatmanlab / pyAFQ

Automated Fiber Quantification ... in Python

http://yeatmanlab.github.io/pyAFQ/

BSD 2-Clause "Simplified" License

57 stars 34 forks source link

Template space for recobundles: ICMB 2009a vs. ICBM 2009c #574

Open jyeatman opened 4 years ago

jyeatman commented 4 years ago

We use ICBM 2009c. Apparently the recobundles atlas is in ICBM 2009a space. I don't know the difference. But this is quoting from the README file that comes with the Atlas_80_Bundles

All the bundles and trk (Trackvis) files are in MNI space ICBM 2009a.

Do contact the DIPY developers at https://gitter.im/nipy/dipy if you need the bundles in ICBM 2009c space.

36000 commented 3 years ago

Nice find. According to : http://www.bic.mni.mcgill.ca/ServicesAtlases/ICBM152NLin2009 It looks like the mean difference between 2009a and 2009c is "sampling is different". Also, "different versions of N3 algorithm produces slightly different tissue probability maps". But they both end up as 1x1x1mm templates.

36000 commented 3 years ago

Also, we use the asymmetric version. I will check if that's what recobundles is using.

Edit: RecoBundles is also using asymmetric.

arokem commented 3 years ago

This exchange from this morning also seems relevant: https://mail.python.org/archives/list/dipy@python.org/thread/MIEWXNAYV5BZD5CFXYPBAGZ5LXV6TFUX/