VIRTUAL REALITY

Envision an emergency room sometime in the future: A patient who has suffered a brain embolism is brought in. The attending physician determines that surgery will be needed to drain the blood and relieve pressure on the brain. However, can all the lost blood be removed quickly, safely and accurately? The doctor orders an MRI scan taken of the patient’s brain, but even this may not be precise enough to ensure there is no permanent damage to the patient. The doctor feeds the scan into a computer, punches a couple of buttons and is rewarded with a complete three-dimensional image of the patient's brain. More keys are punched, and the doctor is able to peel away layers of the virtual image, highlight all the areas of the brain where blood has hemorrhaged. The physician runs a quick walk-through tutorial of exactly what will be needed for the surgery before actually going into the operating room.

For now, this scene is only science fiction, but researchers at Texas Tech University are working to help bring this technology into existence with pioneering work in virtual reality. Bharti Temkin, Ph.D., associate professor of computer science and research professor of surgery, directs the program designed to create three-dimensional computer-based representations of the human body to assist with medical training and to help doctors prepare for surgeries. The process of creating such representations and interacting with them is known as virtual reality, or VR. In her laboratory, medical students and surgeons interact with the simulation using haptic devices or electronic sensing devices that replicate the sense of touch. Touch is critical in performing surgery and being able to “touch” the digital body results in a realistic experience.

"Our work has two important applications," says Temkin. "First, we want to train medical students before they perform the actual surgery. This is similar to training pilots using flight simulators. The use of a computer allows for execution of basic manipulations that can be performed and evaluated by students and teachers before actual surgical procedures are performed. Secondly, we want to give practicing surgeons the opportunity to learn new skills and procedures unfamiliar to them. Overall, this type of system will change the fundamentals of a surgeon’s training."

Eric Acosta, a graduate student working toward his doctorate in computer science, has been involved with the virtual reality project since 1998. He says an important part of creating a VR surgical simulator is making sure the virtual patients are not cookie-cutter copies of each other. One should remember that details of human anatomy vary from person to person.

"What we want to be able to do is present a variety of cases," said Acosta, “something that is not commonly seen. We also can create different virtual patients for the same surgery so one can train for the same surgery, but based on different patients. We also want to be able to evaluate the surgeons and see how well they perform their tasks."

Acosta says virtual reality programs are important in giving students and doctors the ability to "touch" the computer-generated objects.

"Not only do we want to present a visual representation of the virtual world, but we want to be able to give doctors the ability to touch these objects," said Acosta. "In real life, we use the sense of touch to get a lot of information."

Using a variety of haptic devices allows users to touch and interact with objects that only exist in a computer's hard drive. One of the devices of touch resembles a fountain pen attached to a mechanical arm. When moved about, the pen creates a cursor on the computer screen to indicate its position. When the cursor moves against a three-dimensional image on the screen, the pen generates resistance, allowing the user to feel all along the front, top, bottom, sides and even the back of the object. Several different shapes are included in the computer program, from simple geometric shapes, like spheres and cones, to letters of the alphabet, to parts of the body, including kidney and skeletal hand models.

Another haptic device allows users to reach into a virtual patient to clamp and snip blood vessels. The vessels themselves can be lifted, pulled and stretched, but not too far – too much yanking on a virtual vessel will break it. The current program does not allow for any spilling of blood if a vessel is torn, though Acosta said he is considering adding that option.

Temkin notes that virtual reality could allow doctors-in-training to learn important skills that usually must be gained by practicing on live patients.

"Most of the residents observe some surgeries but do not necessarily have first-hand experience," said Temkin. "They may be looking over someone's shoulder while they do the surgery, or perhaps they may participate in a portion of the surgery. The real training often occurs on real patients, and one mistake can be bad. If they can practice on a virtual patient on a computer, they can learn from their mistakes without worrying about harming a real person." Continuing training also is important for practicing physicians and surgeons, who attempt to continue their training.

"Remember, the medical field is not static," Temkin said. "It's a very dynamic field. As new technology comes along, the doctors who are doing surgery now need to be trained for that new technology. They need to practice the new techniques, so if you can train them in a virtual environment, the learning is much faster and the knowledge is safer for the patient."

 Virtual reality also can be used as a diagnostic tool. Acosta has created a program that can take any CT or MRI scan and convert it into three-dimensional, interactive virtual structures. “Using that information, we identify and isolate anatomical structures or organs of interest to create the three-dimensional models,” Tempkin says.

“We want to be able to create a virtual patient," says Acosta. "One of the data sets that are becoming fairly standard is the Visible Human. This is a project from the National Library of Medicine, where researchers froze a cadaver, took CT and MRI scans, took one-millimeter slices from head to toe and came up with 1,876 slices. They took pictures of those slices, then went back and labeled all those images, pixel by pixel, as to what those structures were. Using that information, we can graph any MRI with every part of the body, every structure, every organ."