M.Sc Thesis

M.Sc StudentAlbocher Dan
SubjectReal-Time Haptic Simulation for Surgical Procedures
DepartmentDepartment of Computer Science
Supervisor PROF. Gershon Elber
Full Thesis textFull thesis text - English Version


Virtual reality surgery simulation systems are becoming more and more common nowadays. They can allow medical students as well as expert medical doctors, to train on new surgical procedures while avoiding practice on animals or cadavers, and without endangering patients. For a surgical simulation to be able to assist in this type of training, it must successfully emulate the important aspects of performing a surgery, as they are perceived by the surgeon.

Surgical operations may differ in the way the visual information is conveyed to the surgeon. Endoscopical procedures include the insertion of cameras into the body of the patient, and provide visual feedback via a monitor. In other procedures, the surgeon views the patient directly. Different procedures also make use of different tools. Laparoscopy involves the insertion of a long tube-shaped tool through a small incision at the abdomen, allowing the surgeon to control the laparoscope using a handle at its external end. Plastic surgery, on the other hand, commonly makes use of simple tools such as scalpels, hooks, tweezers, etc. When designing a surgical training simulation system, the simulation hardware and software must be tailored to the type of interaction defined by the simulated procedure.

In this work, we present a simulation system for near-surface surgery that allows full visual and haptic interaction. The operations supported by our system include deformation of the skin and near-surface volume, drawing on the skin while planning the virtual surgery and performing incisions and undermining of the skin, all in real time. The necessary hardware elements for this type of surgery simulation are described, and the software algorithms are presented in detail.

On the software side, a unique prism based mass-spring physical simulation method is proposed, allowing efficient implementation of the necessary operations of near-surface (skin) surgery. Solutions to the specific challenges raised from real time incisioning and undermining with force feedback are also portrayed, along with methods of deforming internal near-surface geometry, such as glands, nerves or blood vessels. Results from our implementation of this system for facial plastic surgery are presented and discussed.