M.U. researchers invent system for predicting falls in the elderly

This is an archived article and the information in the article may be outdated. Please look at the time stamp on the story to see when it was last updated.
Data pix.

COLUMBIA, Mo. -- Falls send nearly three million older Americans to the hospital each year. Their injuries, ranging from broken hips to brain trauma, cost $31 billion to treat. For many, it's the end of their independence. But what if you could predict someone is going to fall and intervene? It's happening in Columbia.

Home is where Ann Gowan's heart is. It's also where the retired gerontologist knows her brain needs to be.

"I have enough to do in the house that my brain keeps going," said Gowan.

Since the 88-year-old fell on her driveway and broke an ankle, she's been more aware that another fall could land her away from home in nursing care.

Dr. Marjorie Skubic, a University of Missouri engineering researcher, saw what happened after her mother-in-law fell.

"Her shoulder never healed properly, so she was in pain for the rest of her life," said Dr. Skubic.

Dr. Marilyn Rantz, a M.U. nursing researcher, saw what happened to her mother.

"She lay on the floor for eight hours and she was dead within six months," said Dr. Rantz.

Necessity may be the mother of invention, but in this case, desire by the researchers to help others led to a system for detecting and even predicting falls. Gowan has motion detectors in every room of her home.

"We also have a depth sensor that gives you a three-dimensional silhouette of the person walking around the environment," said Dr. Skubic.

Dr. Skubic shows an image of Gowan walking into her bedroom. There's nothing Gowan has to wear. The system, without anyone monitoring the images, can detect a fall. An e-mail alert goes to staff at TigerPlace, a retirement residence in Columbia.

"And the video appears...I can see the person actually fell," said Dr. Rantz as she looks at an alert on her mobile phone.

The system can also predict a fall is likely to occur. The depth sensor picks up walking patterns including speed, stride time and length.

"We were able to show that changes in walking speed and stride length over a relatively short period of time are a strong predictor that somebody is going to fall in the next three weeks," said Dr. Skubic, referring to a study of TigerPlace residents.

"I was so stunned I broke into tears because I thought this -- I really could have had this information for my mother," said Dr. Rantz.

When walking changes occur, the system triggers alerts.

"I would have been able to arrange for physical therapy, perhaps get her to balance clinic," said Dr. Rantz.

Possibly preventing a fall.

Gowan also has a sensor in her bed that measures her pulse, breathing rate and restlessness.

"They called once and said, 'Are you okay?' And I said, 'You know, as a matter of fact, I had an infection the other day and I'm not really as well as I had been, and you picked that up?' And they said yes," said Gowan.

The researchers say changes in walking patterns can also be an indicator of health problems in addition to the higher fall risk.

"Our perspective has never been let's figure out when we should kick mom out of her house," said Dr. Skubic.

It's all about keeping her there.

"It's giving me peace of mind. Giving me and my children peace of mind. I think Silicon Valley has their eye right on it. Yea, it's a big one," said Gowan.

Dr. Rantz says the system should be widely available in a year or so and will likely cost more than $200 a month. She expects the price will drop over time.

Tracking Coronavirus

More Tracking Coronavirus



More News