Exploring the idea of bringing together deformable object simulation and plastic surgery is the primary focus of my Ph.D work at the University of Wisconsin. The goals of this project are far reaching, but in short, we aim to provide surgeons a three dimensional authoring environment in which they can craft complex repairs, simulate the potential success of an operation, and record their knowledge of procedures as future resources for education.
As a tool for education and experimentation, it is important that the simulation be responsive to the user. While there may be room for slower, more accurate simulations in prediction oriented tasks, during authoring scenarios the tool must respond in real time to a surgeons actions.
Complicating this requirement is the fact that many common platforms, (cellphones, tablets, most desktops) simply lack the nessessary computational performance to deliver accurate simulations under realtime constraints. As such, this project is also exploring methods of providing simulation as a service, relagating intensive computational workloads to a remote server infrastructure.
It is important that users are not restricted as to how cuts are performed in the virtual tissue. Procedure design is complex and often requires dealing with the geometry of the local skin region on a case by case basis. As such, freeform cutting must be supported to grant the user this needed flexiblity.
However, it is insufficient to merely cut into a geometric model blindly. Anatomical details act as important constraints during an operation, which trained surgeons use in making decisions about their patient's operation. As such, any cutting algorithm or interface to such must take anatomical context into consideration when determining aspects such as depth and angle of any cut.
Correctly capturing the biological details of the human body is essential to surgical simulation. While the obvious details include appropriate anatomical modeling for geometry, there are other, less obvious, considerations that need to be made for plastic surgery simulations.
Along with the geometric extents of the human body, physically accurate material models also must be provided for useful simulation. In plastic surgery, one common and important task is the closing of wounds or holes. Proper limiting of tissue elasticity is essential to determining if a proposed tissue flap created by a surgeon can be successfully used to close such wounds.
Beyond the norm, there are then the patient specific modeling parameters. From extreme geometry due to injury or genetic defects, to wildly varying material parameters due to a host of factors, including age, gender, and ethnicity, creating an accurate model that represents a single patient can be a incredibly complex proposition.