[Insight-users] NMI Vs mattes

[mqutaish]

I used normalizedmutualinformationhistogramimagetoimagemetric and
mattesmutualinformationimagetoimagemetric to rigid register 2 volumes, the
results of NMI were better than the mattes results. i used the same
optimizer with both regular step decent gradient. however, NMI is very very
slow and not threaded as compared to mattes. it is  ridiculous that each
iteration of NMI takes 15-20 min with my data. 

 I searched a lot on this forum but i didn't find a direct answer to how to
multi-thread NMI .
Has anybody tried multi-threading the NMI before? 
some posts suggested modifying some lines in mattes code to get NMI but i
need the one that has the word  histogram as in the above. Can anybody help
me regarding this issue , or even give me direction on how to do the
multi-threading ??

Thanks in advance 




--
View this message in context: http://itk-insight-users.2283740.n2.nabble.com/NMI-Vs-mattes-tp7582300.html
Sent from the ITK Insight Users mailing list archive at Nabble.com.