With this pioneering "Virtual Human Model for Medical Applications", physicians will wear a customised virtual-reality glove during the patient examination that collects data on what the physician is feeling through sensors located in the glove's fingertips. Dr. Thenkurussi Kesavadas, Ph.D., University at Buffalo assistant professor of mechanical and aerospace engineering, director of the university's Virtual Reality Lab and the project's principal investigator, explained that at this time, there is no way that a physician at a second site can share that experience without personally examining the patient. In very serious cases, such as when a patient has been diagnosed at a small, rural hospital, the patient may have to be airlifted to a more comprehensive medical facility where he or she can be examined in person.
The VR system under development at University at Buffalo could make some of those costly, not to mention traumatic, airlifts unnecessary. "Using our customised data-collection glove and the detailed understanding we are developing about the physics behind a doctor's touch during an exam, we expect within two to three years to have a device in use which will allow a physician to apply medical palpation virtually and in real time", stated Dr. Kesavadas. "The system will enhance the current method of clinical palpation by transforming this approach from a qualitative to a quantitative examination", said James Mayrose, research assistant professor in the University at Buffalo Department of Emergency Medicine, doctoral candidate in the UB Department of Mechanical and Aerospace Engineering, a senior designer of the glove, and co-investigator on the project.
Dr. Mayrose and medical professionals in the UB Departments of Emergency Medicine and Radiology are now conducting studies of the glove with human subjects at the Erie County Medical Center. "The UB work represents a departure from the usual route taken by researchers studying VR for use in medical applications", as Dr. Kesavadas noted. "Just about everyone who is looking at virtual medicine for the moment is interested in surgical applications", he commented. Unfortunately however, those applications are many years away. Dr. Kesavadas sees no reason to wait to reap the benefits of VR for diagnostics.
"This system could revolutionise imaging in medicine", according to Anthony Billittier, M.D., medical director for the Office of Prehospital Care at the Erie County Medical Center, and co-director of the Calspan-University at Buffalo Research Center's CenTIR, which is the Center for Transportation Injury Research, and which is funding the work. Dr. Billittier is particularly excited about the UB researchers' creation of a database of information that accurately describes the biomechanical properties of soft tissue under various conditions.
"Right now, if a patient has been in a car crash and has abdominal pain, for example, we can use ultrasound in the trauma room to tell us if there is fluid in the abdomen", explained Dr. Billittier, "but we cannot really tell why. Is it a shattered spleen or a lacerated liver? A database with information in it which could tell us that, just based on the consistency of what the physician is feeling, could allow the surgeons to go right into the operating room without having to obtain a CAT scan. It could save time and, with many injuries, that is absolutely critical."
The University at Buffalo research group is modelling on the computer the soft tissue and organs of the human abdomen, using atomic-unit-type modelling which breaks up human tissues into pieces measuring no more than 8mm. The system takes as its raw material the Visible Human Data Set developed by the National Institutes of Health which features complete, digitised data sets of the human body. Using a very powerful graphics computer, the researchers "supersample" smaller and smaller sections of the data set for a given body part or organ, enabling them to get more and more detailed pictures of each one and develop increasingly complex equations about how each tiny section will respond to applied forces. They then create layers of these sections, gradually building the collection of samples into the complete organ.
"Our big contribution is that we are writing algorithms to model how soft tissue deforms as a real mass, rather than just as a surface, which is what many groups are currently doing. No one else is doing this in real time", stated Kevin Chugh, a UB doctoral student in computer science who is a co-author on the research. "We will be able to touch the model with a haptic thimble, the physically based VR counterpart of a computer mouse, on the screen, apply the force using a haptics feedback system and show how it deforms and then bounces back when the force is withdrawn."
The work is based on a solid understanding of the physics behind what happens when pressure is applied to different parts of the human body. "While the physician is doing a palpation on a patient, the computer, through the VR glove, is picking up all the information about what anatomic-force characteristics the doctor's finger is feeling", said Dr. Kesavadas. He noted that only a handful of groups in the United States are performing atomic-unit modelling for an interactive VR environment. The system will have emergency-services, military and battlefield applications.
University at Buffalo's Virtual Human Model also could enhance the training of new physicians in one of the most formidable medical procedures they have to learn and train, being intubation. "Intubation is a very touchy procedure", Dr. Billittier explained. "It involves putting a plastic tube into a patient's airway or windpipe whenever that patient may not be breathing or may be in shock." He pointed out that medical students and emergency-medical technicians currently learn the mechanics of the procedure first on a mannequin, which is not very lifelike, and then switch to real patients, with widely varying degrees of success, under the close supervision of an experienced physician.
A computerised VR simulator for intubation based on information being gathered for the Virtual Human Model database would be a real boon, according to Dr. Billittier, who is working with Dr. Kesavadas' group to develop one. Such a system could provide visual feedback of real airways while the student is manoeuvring the tube into the simulated patient on the screen. The simulation would provide the student with the proper tactile or pressure feedback, as it is felt through the haptic interface. Dr. Kesavadas estimated the research team could develop such a system within a couple of years.