AISIM Task Repartition

---

index :

(Bio)Medical teams

Computer Graphics teams

Conclusions

[Back to AISIM]

During our first meeting (12-13 December 1997), we decide the following task repartition:

(Bio)Medical teams

IRCAD	M3N/Mostra
Description and geometrical model of the surgeon's tools Surgery description(video and comments) Role and behavior of organs in the vicinity of the liver	Biomechanical reference model of the liver Biomechanical validation Reference deformations
Medical glossary Liver anatomy: structure position in the body anchor points

Computer Graphics teams

Sinus/Epidaure	iMAGIS	Sharp
Geometrical models of liver's neighbors Optimisation and parallel computation of deformable models	Physical models of liver's neighbors (behavior interactions with the liver) Realistic rendering (specularity to improve 3D perception) Textures : organs surface (3D perception) inside the organ (bleeding) aspect modification due to tools' action aspect modification due to pathology
Real time deformable models cutting and suturing taking into account movments due to breathing heterogeneous model (vessels, pathology) contact with tool (force feedback, friction) Vessels model hepathic pedicle cystic duc contact between several tools (force feedback)

Conclusions

AISIM's members will start by focussing on minimal invasive techniques for liver surgery. Then, the brain surgery case could be studied.

For the part common to Epidaure/Sinus, iMAGIS and Sharp, we plan to do as follow :

Concurrently, the different teams will work on their own deformable model. These models will be compared amongst themselves and to the biomechanic reference model from M3N/MOSTRA. This evaluation will focus on :
• deformation quality (as similar as possible to the reference model)
• computation time (real time)
• the ability to cut and suture the model in real time
We shall not forget the rendering aspect (texture mapping).

Different teams are also encouraged to develop bilateral relations among themselves.
Each team is asked to provide as soon as possible a preliminary program and a detailed presentation of their research.

Some questions remain opened, mainly about haptic simulation:

• Do we really need a very high frequency (>300Hz) update of exerted forces? Can we accept a low frequency update (with interpolation)?
• What is the infulence of the tools speed on this problem?
• Taking into account high friction due to the tools themselves, what are negligible phenomena? What are unavoidable phenomena?