While older adults aren’t the target audience for new video game technology, University of Missouri researchers are using the devices to help keep seniors healthy. Marjorie Skubic, a professor of electrical and computer engineering, teamed up with TigerPlace, an independent living community, and experts in nursing, social work, medicine, health management and other fields to develop unique monitoring techniques for seniors. Below are excerpts from our recent interview.
For years, you used motion-sensing technology to monitor changes to the health of TigerPlace residents. What did that entail?
We’ve been working on this idea of monitoring elderly people in their homes for about seven years. We’ve had motion sensors installed in TigerPlace since 2005. The main idea is to do early detection of illnesses and functional decline. We’re looking for a proactive healthcare model. You can prevent a small problem from becoming a catastrophic problem.
We’ve detected urinary tract infections. That’s a simple fix if you catch it early. We’ve been able to show with our sensors that we can identify changes that correlated with things like UTIs before they were caught using traditional methods.
How did these sensors work?
All the sensors we’re using are environmentally-mounted. We don’t use any wearable sensors. We want to capture the normal daily activities of a resident. We don’t want them to have to worry about putting something on. It’s completely transparent. We’re passively observing what they’re doing. The motion sensors are mounted on the walls. We also put them in specific places to capture certain kinds of activities. The sensors will [capture] motion about every seven seconds if there is continuous motion in an environment. If you process that as a density, you can get an indication of how active an individual is. You can tell whether someone is more sedentary. That’s part of the way we use the data.
Now, you’re using Kinect, a motion-sensing camera used as a video gaming device, to monitor behavior and changes in patients. How does the camera work and what specifically are you monitoring?
Continuing with this idea of early illness detection, another indication of declining health is a person’s gait pattern — how somebody walks in terms of speed, step length, how they move back and forth. These can become a sign of health problems. A lot of people are concerned about falls. We have two grants funded by different agencies. Both are centered on the idea of fall detection and fall risk assessment. We’re using a variety of sensing mechanisms to detect falls and measure things like gait patterns. You can’t get this from passive sensors. That’s why we started looking at other sensing modalities.
We started looking at the Kinect when it became available last fall. A two-camera system extracts silhouettes of people as they move around. The silhouettes are for privacy protection. From that, we create a three-dimensional model, so you can capture how someone walks. One nice thing about the Kinect is it does not require visible light. It works in the dark and low-lighting conditions. It’s an interesting platform in which to explore this. It’s relatively inexpensive.
Another project is a fall detection system that uses Doppler radar to recognize changes in walking, bending and movements that might indicate heightened risk for falls. Talk more about this.
We’re looking at whether we can get gait information from the radar. This is a collaborative project with GE Global Research. We’re using a radar unit that had been used in security systems. It uses the principles of Doppler radar, which means it’s detecting velocity. We’re looking for energy bursts from the radar that could be an indication of falls. It sees through structures, so you don’t have the problem of furniture obstructing sight.
What happens if the systems sense a problem for a patient? Where is the data sent?
The data goes into a server. There’s a web interface where the clinical staff can look at the data. It’s a lot of data. It’s not practical for someone to sit there and look at it. We instituted an automated alert. The sensor system is not trying to diagnose the problem. It’s trying to let the clinicians know something is going on and they should take a closer look. The alerts are emails to them. In the email is a link to the web-based sensor interface that pops up a window that gives them the alert in the context of what’s going on. It’s easy for them to quickly look at the data and different parameters. We’ve got that all set up.
What’s the next step for this work?
The early illness model with the motion and bed sensors could be extended now beyond TigerPlace. We’re looking for opportunities to do that. In order to show the efficacy of this, it’d have to be done on a larger scale. I think we’ll get there. We have several people we’re talking to about how we might scale up.
Everything we’re doing there could be done in a way that it could be implemented in the private home. We’d like people to stay healthy and safe and independent in their homes or wherever they choose. We want to give people the freedom to age in whatever place they want. That’s where we’re headed with this — to try to keep people in their own homes.
Photo: Marjorie Skubic