Changes for page 4. Registration pipeline

Last modified by rbakker on 2023/07/17 12:36

From version 2.1
edited by rbakker
on 2023/07/17 12:01
Change comment: There is no comment for this version
To version 1.1
edited by rbakker
on 2023/07/17 11:59
Change comment: There is no comment for this version

Summary

Details

Page properties
Content
... ... @@ -3,4 +3,4 @@
3 3  
4 4  After the slices have been aligned and the pieces of neurite have been stitched together in a single valid neuron, the neuron registration pipeline is used to warp the neuron to a common reference space. The strategy to follow depends a lot on the usecase, in particular on whether the slices have been badly deformed in shape or contrast.
5 5  
6 -In the collab [[morphology-atlas-registration-v2>>https://wiki.ebrains.eu/bin/view/Collabs/morphology-atlas-registration-v2]] we present a usecase in which the sections of a neuron are not good enough to be automatically registered into a 3d volume. Instead, each section is first roughly registered to the Allen Mouse refertence space CCF3 using [[QuickNii>>doc:Collabs.quicknii-and-visualign.WebHome]], and then the [[Elastix package>>https://elastix.lumc.nl/]] is used for additional nonlinear warping. Registering each section individually does come at a price: the resulting neuron has discontinuities. These are resolved by finding the best possible linear transform between the original neuron and the registered neuron. This linearly warped neuron is then used as the end result.
6 +In the collab [[morphology-atlas-registration-v2>>https://wiki.ebrains.eu/bin/view/Collabs/morphology-atlas-registration-v2]] we present a usecase in which the sections of a neuron cannot be automatically registered into a volume. Instead, each section is first roughly registered to the Allen Mouse refertence space CCF3 using [[QuickNii>>doc:Collabs.quicknii-and-visualign.WebHome]], and then the [[Elastix package>>https://elastix.lumc.nl/]] is used for additional nonlinear warping. Registering each section individually does come at a price: the resulting neuron has discontinuities. These are resolved by finding the best possible linear transform between the original neuron and the registered neuron. This linearly warped neuron is then used as the end result.