A Dublin, Ireland-based company, Kinesis, has found that its QTUG system, which uses app-connected wearable sensors to assess fall risk, could be valuable for evaluating multiple sclerosis, too. Studies have found the technology to be 10 to 20 percent more accurate than existing methods for detecting falls.
Kinesis, a spin-off of Dublin’s TRIL Centre for aging technology research, launched last summer after collecting validation data for more than seven years. In October, they signed a distribution agreement with Intel-GE Care Innovations to bring the system, which is FDA registered as a Class 1 device, to the US. Intel and GE also help fund the TRIL Centre.
“It’s based on a standard clinical test call the Timed Up and Go which is used throughout the world to assess mobility and fall risk,” Dr. Barry Greene, Kinesis’s Chief Technology Officer, told MobiHealthNews. “What we do is instrument this test using body worn sensors placed on the legs. So the Timed Up and Go is a test where a subject gets out of a chair, walks three meters, turns around, walks back and then sits back down again. … We’ve come up with a statistical algorithm based on the thousands of assessments we’ve done to come up a probabilistic estimate of the risk of falls, but also identify what element of the patient’s mobility is creating the risk of falls. Then all that data is streamed via Bluetooth to an Android tablet, which then displays the information to a clinician.”
The studies compared the QTUG (which stands for Quantified Timed Up and Go) to some popular existing methods for assessing fall risk, including a non-quantified TUG test, where doctors just observe patients moving and time them with a stop watch, and grip assessment. Other popular methods are little more than questionnaires, Greene said.
The recent MS research, presented at the Engineering in Medicine and Biology Society (EMBC), 2014 International Conference of the IEEE, showed that 32 of the 52 sensor parameters used by the QTUG could reliably assess disease state in 21 patients with relapsing remitting multiple sclerosis. Previous research established that the QTUG data could distinguish between MS patients and controls, whereas the stopwatch version of the TUG test could not.
“We have a large patent portfolio,” Greene said. “What we started out with was looking at this for falls, but we’re also very interested in looking at this for neurological disorders. So we’ve just completed a study on MS and looking at using this for assessment of disease state in MS and there’s also potential for use of this technology in Parkinson’s disease, looking at the response to medication.”
In a 2010 study in IEEE Transactions on Biomedical Engineering, the manual time TUG was found to have a sensitivity of 58 percent, compared to a QTUG sensitivity of 77 percent. Another popular assessment tool, the Berg Balance Scale, had a sensitivity of 65 percent. For specificity, the QTUG came in at 76 percent compared to 58 percent from the TUG test and 64 percent for the Berg Balance Scale.
UPDATE: Added links and names of journals where the two studies were published.
While some advocate at-home tracking with devices like the Fitbit as a way to get a more complete picture of a patient’s movement, Greene says when it comes to fall assessment just the opposite is true: data is most valuable when researchers can control the context around it, as they can in a clinical assessment setting.
“I was involved in the development of a study a few years ago where we put sensors in the home of 10 elderly people and we kind of measured their gait speed as they walked around,” he said. “What the project essentially floundered on was we had no way of knowing if they were walking slow because they were reading the paper or if they were walking slow because they were ill. I think that’s a fundamental problem that a lot of these wearables have is that there is no context and it’s very difficult to make it clinically meaningful.”
