registration time is too long? about antspy

[stnava]

Yes. You might want to adjust the parameters if you were concerned about speed.

On Fri, Jun 28, 2019 at 1:23 AM chestnut111 ***@***.***> wrote: hello，I install the antspy by pip install git+https://github.com/ANTsX/ANTsPy.git I would like to use ants to register on MRI data of chest。 my code ： out = ants.registration(fixed, moving, mask=mask, type_of_transform='SyNCC') the data: [image: image] <https://user-images.githubusercontent.com/26037045/60319044-be220700-99a7-11e9-83da-ad9e512b3246.png> Is this normal that use 3 hours to run this code ?? thx !! — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#62?email_source=notifications&email_token=AACPE7VAS6QVN4ZTI7YWNXLP4WN33A5CNFSM4H4CG322YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4G4HODJQ>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AACPE7V5ITAOFKMRCDVJSUTP4WN33ANCNFSM4H4CG32Q> .

-- brian

[stnava]

For example, use SyN.

[fconstance]

Hello,
Thank you very much for the great code, the registration works very well.
I have a question about the computation of the MI and NCC metrics when performing registration with SyN or SyNCC transforms. Why is NCC so much longer?
Thank you very much for your answer,
Constance

[stnava]

CC has much more to compute. you can read the papers: https://pubmed.ncbi.nlm.nih.gov/20851191/

but in brief:

• CC is a local metric and thus has "special" computation at each voxel
• MI is a global metric and has relatively little special work to do at each voxel

[fconstance]

Okay, thank you very much for the fast answer and the paper !