TechWeb

Microsoft Kinect Technology Helps Surgeons Manipulate Images

Jan 18, 2013 (05:01 AM EST)

Read the Original Article at http://www.informationweek.com/news/showArticle.jhtml?articleID=240146565


10 Medical Robots That Could Change Healthcare
10 Medical Robots That Could Change Healthcare
(click image for larger view and for slideshow)
A system based partly on the Xbox Kinect motion-sensing technology can be effective in helping surgeons manipulate MR images in the OR while preserving sterility, according to a new study in the Journal of the American Medical Informatics Association.

Surgeons are interested in this technology because the manual use of computer workstations to manipulate these images creates potential for infection. Physicians who touch a mouse or a keyboard need to scrub in again, wasting precious time. Although surgeons can ask staffers to rotate or enlarge images, this approach can be cumbersome, inefficient and frustrating.

The hands-free Kinect gaming device, which Microsoft created in 2010, has been used before to safeguard sterility in ORs. For example, surgeons at Sunnybrook Hospital in Toronto have been using Kinect since 2011 to manipulate images without touching a computer. But the JAMIA study, conducted by researchers at Purdue University in Indiana, used scientific methods to establish the reliability and efficacy of the technique.

[ Technology can play a key role in improving health care, but it will never replace the human touch. Read Internet Has Yet To Replace Family Doctors. ]

The team used a three-step process in their research. First, they developed an MR image browser application and studied the gestures of 10 surgeons to establish which of their gestures would be appropriate to use with the application. Second, the team developed gesture-recognition software that would detect the intentionality of users' gestures. And finally, they performed three experiments to validate whether the system could discriminate between intentional and unintentional gestures.

The first experiment, which included both intentional and unintentional gestures, showed that the system could detect the correct gestures with an accuracy rate of 97.9%, with a 1.36% false positive rate. Similar results were obtained when the number of users was increased from two to ten in the second experiment. In the third experiment, the research team determined that the system could correctly determine intent or context 98.7% of the time. Mean gesture recognition accuracy was 92.6% with context and 93.6% without context.

Juan Pablo Wachs, an assistant professor at Purdue University and one of the study's authors, told InformationWeek Healthcare that the importance of context was the key finding in the study. "When you don't look at the context, you're just looking at the hand movement to control a medical image in a browser. There are a lot of reasons a surgeon may want to perform hand movements. The critical point is that the application has to be able to recognize the intent of the surgeon to be able to know whether he's performing movement to manipulate the images in the browser."

The high rate of gesture recognition using motion-sensing technology is not new, Wachs pointed out. What is new is that the system his team created decreased the number of "false alarms" -- gestures that were unintentional or that had nothing to do with image manipulation.

Wachs added that he and his colleagues have been studying the use of motion-sensing devices in the OR since 2008, long before Kinect came along. "Kinect helped us get better results, but it was not solving the scientific problems. It just provided a tool."

Despite the promising results of this study, Wachs and his colleagues conclude that more studies are needed with larger groups of surgeons, since people gesture in different ways and signal their intent to use the system differently.

Clinical, patient engagement, and consumer apps promise to re-energize healthcare. Also in the new, all-digital Mobile Power issue of InformationWeek Healthcare: Comparative effectiveness research taps the IT toolbox to compare treatments to determine which ones are most effective. (Free registration required.)