Building a "flight simulator" for surgeons: Project joins computer science with medicine

A UW surgical resident views a simulation on a computer screen

"You need to screw up.  That’s how you learn,” says Dr. Court Cutting, a world-renowned expert in cleft lip and palate procedures, to a roomful of UW plastic surgery residents early on an October morning.  “But you should screw up there, on a simulator,” he continues, gesturing to a computer monitor.  In a crowded UW Hospital conference room, the residents—just beginning their surgical careers—take in Cutting’s candid advice, formed through decades of experience.

There’s just one catch:  right now, there isn’t a highly sophisticated, real-time simulator for surgical procedures.  But assistant professor of computer sciences Eftychios Sifakis is working hard to change that.

Sifakis, collaborating with Cutting, Dr. Timothy King and computer sciences graduate student Nathan Mitchell, is developing a “Skin Biophysics Surgical Simulator” that will help surgeons-in-training model procedures in detail before operating on live patients.

Cutting likens it to flight simulators that give novice pilots extensive training before they enter real cockpits.  “You can work out concepts and get well developed on how to handle almost any [flight] situation,” says Cutting, citing examples like coping with engine failure or malfunctioning landing gear.  “Why don’t we train a surgeon in a simulator?  Because there aren’t any.”

The National Science Foundation is funding the team’s efforts through its “Smart and Connected Health” initiative.  Sifakis is serving as principal investigator for a three-year grant awarded in July 2014 (King, of the UW-Madison School of Medicine and Public Health, will serve as co-PI).  This work has immense potential to reduce errors, minimize post-surgical complications and improve patients’ quality of life.

Development of the simulator has already begun.  During the Oct. 3 meeting with surgical residents, Sifakis, Cutting and King gave them a first glimpse at what the team has been working on.  Cutting demonstrated ways to make and close incisions on flat areas of skin, as well as make scalp incisions and suture them through flap techniques that promote the best possible healing.

Biomechanical accuracy is what makes Sifakis’ project important.  Ultimately, the three-dimensional renderings produced by the simulator will demonstrate how real skin will react under real conditions.  Though this requires further development—as well as the acquisition of more data about live tissues—this approach has the potential to revolutionize surgical training.

“The applications are limitless,” Cutting stresses.  “It’s all of surgery.  The human body is a complex, elastic object, and Eftychios Sifakis is going to be one of the first developing realistic, bioelastic models of the body.  He’s a pioneer in bioelasticity; it’s never really been modeled like it should be.”

Sifakis looks forward to growing a two-way partnership with the medical community.  “We want to get it [the simulation tool] in their hands and get feedback from them,” says Sifakis, an expert in computer graphics and simulations whose experience ranges from Department of Defense projects to consulting for major companies like Disney Animation Studios.

Sifakis is hungry for “a continuous flow of information” between computer scientists and clinicians.  Their likes and dislikes will guide the simulation tool’s ongoing development.

Indeed, at the October training session, residents were already brainstorming features they’d like to see.  For example, one suggested the capability to superimpose incisions made at different angles, so a surgeon could see how those incisions would create different end results.

“There’s not just one solution to a particular case,” explains Sifakis.  “Different surgeons have their own styles.”  

As the system undergoes further development, accurately representing bioelasticity is crucial.  Surgeons must predict how soft tissue is going to respond to a procedure, and the tensile strength of tissue varies by the type of tissue and its location on the body, as well as by a patient’s age (an 80-year-old’s skin, for example, has vastly different bioelastic properties than a baby’s).

Advances in computing technology have also broadened what is possible.  Sifakis’ end goal is to create a system that can be deployed on a tablet or other mobile device, making it nimble and portable for surgeons.  The high-level computing needed to generate complex simulations will happen remotely on UW’s own servers, then the simulation is sent back to the tablet.

It will likely be several years before Sifakis’ system could be in widespread use, given the need for greater physical accuracy, higher resolution and a diversity of surgical operations.  However, Sifakis and his team are off to a promising start, and future training sessions are planned that will tackle specific procedures, such as breast reconstruction following cancer surgery.

Sifakis is glad to have Cutting, now retired as director of the Cleft Lip and Palate Program at New York University Medical Center, as a collaborator.  "We share a passion for computer modeling of surgery."

[ Photo credit:  On Oct. 3, 2014, a University of Wisconsin surgery resident takes a first look at the surgical simulator under development.  Photo by Sarah K. Morton, UW-Madison College of Letters & Science ]