Virtual reality project shows promise
By Peter Pollack
Thursday marked the 10-year anniversary of the beginnings of the Virtual Reality Arthroscopic Knee Surgery Simulator Project (VRAKSSP), an ambitious attempt to create a paradigm change in the way orthopaedic training is taught in the United States. The project uses modern computer and electromechanical technology to place the student in a simulated working environment so that he or she can perform an arthroscopy on a realistic virtual knee.
On Feb. 15, 1997, W. Dilworth Cannon, MD, hosted a dinner entitled “An Evening of Virtual Reality,” at the AAOS Annual Meeting in San Francisco. Dr. Cannon recalls that the dinner was resoundingly popular, drawing interest from far more people than what may have been expected based on the somewhat primitive forms of virtual reality (VR) possible with the best computers of the era.
“Such a dinner would have normally drawn about 250 people, but we drew close to 900 or 1000,” he says. “They were actually scalping tickets outside the Hilton Ballroom.”
Although that early demonstration of VR’s possibilities was well-attended, it would take several years to acquire the necessary funding and sketch out the earliest design of a working device. It also took time for technology to catch up with the desires of the designers. Dr. Cannon, now the chair of the project’s Content Development Group, holds his arms well apart to describe the size of the computer they used in the first demonstration. Today’s model is a standard tower computer that wouldn’t look out of place in any office.
New technology, new training
The obvious advantages of students honing their techniques in an environment where they can do no harm have led a number of companies and academic institutions to explore the idea of surgical simulators. Simulab Corp. was founded in 1994 to do exactly that, and Stanford University recently invested in a $4-million center that uses mannequins in an operating room setting to bridge the gap between classroom and the real world. Yet most previous attempts at creating a virtual surgery environment have either been simplified variations on real surgery, or targeted to areas outside the specific needs of orthopaedic surgeons.
“Residents follow a basic apprenticeship training, first watching the attending do the surgery, then actually performing parts of the case once they achieve some proficiency,” says Dr. Cannon. “That’s a very slow process; the operating room time goes way up, and surgical errors are another factor. What we hope to do is enable residents to train and reach an acceptable proficiency level before they go into the operating room to do arthroscopic surgery on real cases.”
To produce the working design, Dr. Cannon and his team partnered with Touch of Life Technologies (ToLTech), best known for its work on the Visible Human Project® (VHP)—an initiative undertaken by the National Institutes of Health to create complete, detailed, three-dimensional models of normal male and female bodies. To achieve that goal, VHP sectioned the male human body into 1 mm intervals, and the female human body into 1/3 mm intervals.
Although the degree of resolution attained in the VHP was considerable, the AAOS team wanted the simulation to have an even higher level of realism. Advances in technology gave the ToLTech team the ability to section the knee into 0.1 mm intervals, allowing them to produce a 3-D model with 100 times the resolution of the male specimen in the previous project.
Creating the interface
Once the software model was in place, the next step was to build a unit that could serve as the interface—a set of physical tools that would work and respond like the surgical instruments and knee in a real procedure. Fortunately, available technology had kept pace with the needs of the development team. Instead of being forced to design a complete mechanical interface from scratch, the group was able to procure a pair of haptic devices “off the shelf” from SensAble Technologies of Woburn, Mass.
The key to a haptic device is that it can be tied into software to provide tactile feedback to the operator. In other words, if the probe on the virtual reality unit is used to palpate the lateral meniscus, the user will “feel” the cartilage as though he or she were actually performing the procedure. This allows the resident to develop a touch for the surgery that previously couldn’t be learned without a live patient or a cadaver
The haptic devices installed on the VRAKSSP simulator are capable of six degrees of freedom, so they can handle the full range of motion that would be expected of a surgical instrument. Additionally, the simulator has a model of a leg that can be rotated and manipulated to simulate working with a real patient. The unit can be height-adjusted for a comfortable working environment, can function as either a left knee or a right knee, and is programmable to simulate a variety of pathologies.
A virtual teacher
Perhaps just as integral to the simulator as the high-resolution graphics and the mechanical design is a special program called the mentor. The mentor oversees the student’s progress, while simultaneously offering advice and critique. In a small way, the mentor takes the place of a human teacher, enabling residents to practice with the unit on their own without taking up time in an operating room. The mentor program also functions as a gateway, allowing students to progress to a new technique only after they are deemed proficient at the current one.
Does this imply that someday, surgeons will be trained primarily by computer? Karl Reinig, PhD, of ToLTech believes that the experience and intelligence of a human trainer would be difficult to replace, but he is quick to point out how the simulator can assist physicians and residents alike by providing precision feedback.
“We do have some advantages in being able to be quite objective,” he explains. “As an expert standing over the shoulder, you’re kind of guessing how much force I’m putting on that cartilage, until you see the little line of cartilage torn up by it. The simulator shows exactly how much force you’re putting on at all times, and we can actually get more quantitative. That’s the purpose of the mentor.”
To design the mentor, the development team plans to measure the proficiency of 10 surgeons working in the orthopaedic community. Those measurements will be averaged, and residents will be required to meet that mean level of proficiency to advance. The question of what to measure and how to measure it was left to members of AAOS.
“The rubric that the AAOS task force put together said ‘These are the important points and this is how we score them,’” says Dr. Reinig. “These can be very much quantified and put into the computer. Did they look here? Did they feel this? What conclusions did they draw? The mentor is an eavesdropper—it’s always got access to the things that are going on in the simulator. And it’s a two-way communication: the mentor tells the simulator what to display, and the simulator goes back to the mentor and says ‘These are the things you asked to know about, now you do your grading accordingly.’”
Bringing it to the public
In building such a complex device, the design team still faces a few hurdles before it is prepared to send the simulator out into the world. Still, everyone involved admits that the progress is moving along nicely, and feels that they are on a pace to deliver it for validation trials by the end of 2007.
Validation will take place at eight residency programs across the United States, each of which will be given the chance to use the simulator for 45-to-60 days. Third-year residents will be divided into test and control groups. The control groups will train the way they always have, while the test groups train on the simulator. Eventually, each resident will be asked to perform a diagnostic arthroscopy, and internal and external recorders will be used to record the procedures. A group of physicians will rate the students according to a set of standards, and the results of both groups will be compared.
According to Howard Mevis, AAOS director of electronic media, evaluation and course operations, this study will be the largest of its kind ever carried out in the area of surgical simulators.
Assuming the VRAKSSP unit passes validation, the development team has plans to expand the concept to other joints and other pathologies. In just a few years, residents across the United States may be learning complex knee, shoulder, and elbow techniques and more through the wonders of virtual reality.