Forums Anyscript.org  

Go Back   Forums Anyscript.org > Professional Forum > Debug model

Notices

Reply
 
Thread Tools Display Modes
  #31  
Old 05-18-2017, 12:30 PM
pgalibarov's Avatar
pgalibarov pgalibarov is offline
Senior Member
 
Join Date: Oct 2010
Posts: 926
Default

Hi Zach,

ScaleMat will only be used when you do not have subject-specific scaling of the foot. So not sure why you want it. The #ifndef statement will exclude it in case of the subject-specific scaling.

Kind regards,
Pavel
Reply With Quote
  #32  
Old 08-15-2017, 12:25 PM
Zach_Welshman Zach_Welshman is offline
Member
 
Join Date: Sep 2016
Posts: 38
Default Statistical shape/appearance model integration with the AMS

Hi Pavel, it's been some time since I posted on this thread and I would just like to thank you for your help to date.

I'm not sure if this should be a new thread, but I am keeping it in the same thread so it is easier to find.

I have a question related to the RBF scaling and integration of an active appearance model.

I have an active appearance model of the foot and from this I have .pts files and .cnn files for points in connectivity of quite a few bones in the foot - the shape model itself is quite revealing, but am interested in how I can utilise outputs from the shape model in the AMS

I have exported the AMMR GM foot STL files as my mean shape and built the shape model that way. Although the reference frames between the shape model and AMMR GM are different. I was wondering:

1. Is it possible to switch the low resolution GM stls for the high resolution stl's i have made from winding the .pts and .cnn files from the appearance model?

2. If it is possible to batch process the Inverse dynamic modelling in the AMS of all these patient specific scaled tibia to navicular bones with just a few control points given there was a global registration to the AnyBody GM STL's so all vertices and faces in the shape model correspond.

I hope this makes sense and any thoughts would be greatly recieved.

Many Thanks,
Zach

Last edited by Zach_Welshman; 08-15-2017 at 12:25 PM. Reason: clarification
Reply With Quote
  #33  
Old 08-15-2017, 01:02 PM
pgalibarov's Avatar
pgalibarov pgalibarov is offline
Senior Member
 
Join Date: Oct 2010
Posts: 926
Default

Hi Zach,

Yes, this is all possible. Practically you just need to organize your data and scaling files, so that you can switch between them through a define with the subject ID.

And for the scaling - you need to set up a template scaling, where you will have to make your own source bones from the statistical shape model. So here is a workflow how i see it:
a) separate bones from the GM stl in Meshlab or elsewhere
b) make new STL from the statistical shape model, which would match a) - it can be achieved by optimizing the input parameters with the objective function being a similarity measure between objects, for example. This will enable using corresponding vertices as landmarks in RBF scaling.
c) setting up a combined scaling function to take into account all individual scaling functions.
d) set up batch-processing so that input files will be automatically switched together with the subject id. For example, the filenames could contain subject-id embedded.

Hope this helps.

Kind regards,
Pavel
Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +2. The time now is 12:56 PM.