Merging sulcal segments with AIMS results in inconsistent features
Posted: Thu Mar 23, 2017 9:10 pm
Hello,
I am trying to merge a few nodes in cortical graph with AIMS. Here is the procedure I am following:
1. merge mesh from the segments (using aims.SurfaceManip.meshMerge)
2. assign the merged mesh to one of the nodes (aims.GraphManip.storeAims) and remove all other nodes (graph.removeVertex)
3. run 'Compute Cortical Fold Graph Upgrade From Old' process in order to recalculate morphometric attributes for the merged node from the new, merged, mesh (with the 'compute_fold_meshes' flag set to False).
The 2 first steps seem to work, but the 3rd one did not update the merged node as expected. Looking at the merged node and comparing the values before and after the procedure run, I've seen that:
- some attributes were not updated, i. e. the same value before and after the merging (refsize or refmaxdepth, for example),
- some attributes got higher value, but new value is significantly lower then expected, i.e. significantly lower then a sum of, or even then a single max value across the merged nodes (white_surface_area, for example), and
- some attributes value even went down (refsurface_area, for example)
Is there anything wrong in the procedure I am using? Am I missing anything? Or may be there is a different - better - way to properly merge cortical segments? Any help would be highly appreciated.
Regards,
Ruslana.
I am trying to merge a few nodes in cortical graph with AIMS. Here is the procedure I am following:
1. merge mesh from the segments (using aims.SurfaceManip.meshMerge)
2. assign the merged mesh to one of the nodes (aims.GraphManip.storeAims) and remove all other nodes (graph.removeVertex)
3. run 'Compute Cortical Fold Graph Upgrade From Old' process in order to recalculate morphometric attributes for the merged node from the new, merged, mesh (with the 'compute_fold_meshes' flag set to False).
The 2 first steps seem to work, but the 3rd one did not update the merged node as expected. Looking at the merged node and comparing the values before and after the procedure run, I've seen that:
- some attributes were not updated, i. e. the same value before and after the merging (refsize or refmaxdepth, for example),
- some attributes got higher value, but new value is significantly lower then expected, i.e. significantly lower then a sum of, or even then a single max value across the merged nodes (white_surface_area, for example), and
- some attributes value even went down (refsurface_area, for example)
Is there anything wrong in the procedure I am using? Am I missing anything? Or may be there is a different - better - way to properly merge cortical segments? Any help would be highly appreciated.
Regards,
Ruslana.