TWO .stl file alignment and measurement

Support
,
43

Hello Prof Andras

I hope you’ve been well. I would love to hear your opinion & advice on the direction of our project.

My research team is working on a project where we compare 2 3D scans and essentially compare the % match between the scans. This would be useful for forensic odontology application in identifying victims following a mass disaster event.

The comparison would be between a 3D CBCT scan of the dentition prior to death (from existing dental clinic patient record file) and a 3D CBCT scan of the deceased person.

We are thinking of using the ALPACA’s align functions for this. I would like to ask 2 questions

  1. Is this a feasible idea? I understand that ALPACA’s align functions are near flawless for very similar 3D files, however in this case, some time would’ve passed between the pre-death scan and the post-death scan, meaning some teeth could have drifted, worn away, or had fillings work done, etc. also, keeping in mind only a portion of the dentition may be available for post-death 3D scan (considering there may have been a massive trauma)

  2. Our research team has a rather flexible budget, probably ranging anywhere from $5k - $25k. Would the commercial partners be available to hop on board with a budget like this?

Thanks so much for your help

44

If meshes are very different then you may need to simplify the problem with manual preprocessing (cut off parts of the mesh that are changed; for example using Dynamic Modeler module’s Select by points tool) and manual prealignment (using landmark registration).

45

Thank you Prof Lasso for your insight.

We thought that ALPACA would be a great tool for our project since the whole process seems quick and seamless without the need for manual preprocessing & pre-alignment. Would this still be a mandatory step in our case considering there could be minor/major differences between our samples? Would there be a way to automate this process?

As if a pair is a match, in theory the meshes should be nearly identical.
Your insight is appreciated

46

If ALPACA gives good results without any preparation, cleanup, preprocessing of the data then it’s great. If the method struggles to give accurate results reliably, then you can simplify the problem as I described above.

I think this amount should be sufficient to get help with developing a custom application. I would encourage you to contact Slicer commercial partners.