Rigid simulation of personalized surgical guide placement

Hi. I contributed to the BoneReconstructionPlanner extension of Slicer used to do Virtual Surgical Planning and Patient-specific Surgical Guides of the mandible reconstruction surgery.

Due to some videos I have seen of the makers of the Jaw-in-a-day procedure, where they replace part of the mandible with fibula pieces and put on them dental implants and then a prosthesis, I have the thesis that the guideBase surface (the part of the surgical guide that touches the bone or periosteum) could be relaxed (smoothed or even replaced with a biplanar shape).

To test this, and knowing I have the 3D models of all the possible configurations of surgicalGuides I would like to simulate the placement of the surgical guide over the fibula. Also evaluate stability of the positioning due to random small forces added to the imperfect movement of the surgeon’s hand while placing it. And investigate what guideBase configuration gives the best results. The results calculated as the deviation between the ideal cut plane frame and the final simulated cut frame.

How difficult is to achieve this? How many weeks of work would take from building this simulation library to being able to test the positioning of n surgical guides configuration and do this test on m fibulas. How long computation would take?

Thank you very much


So you have a rigid guide that goes over the mandible? You then want to place it over the mandible and evaluate positional stability? There’s probably a more geometrical solution here rather than physical to precisely evaluate the degrees of freedom the cover has once placed on the mandible (ie: which ways could the shape move without touching).

That being said it is certainly possible to simulate it physically. With one being able to apply various loads in different spots to check the stability of a piece on the mandible.

The easiest way to do it would be to:

  • Setup a non-moving/static signed distance field of the mandible.
  • Setup a rigid body pointset of the piece.
  • Setup collision between them (point based collision). Note the limitation of point based collisions. Contacting edges are missed.
  • Setup the movement (placement) of the piece on the mandible.
  • Evaluate what roll friction would play
  • Perform whatever metrics you’d like to evaluate stability, after or during placement.

For movement you could record it with a haptic device, VR controller, or model it in a 3d program. I would suggest movement with virtual coupling. Such that you are moving the piece with forces rather than directly modifying position. So that when it comes into contact the forces cancel out (whilst the model is free to control positions of things).

1 Like