PORTLAND, Ore.—The infamous 1980s television commercial that featured the tag line "I've fallen and I can't get up" became the butt of a thousand jokes. But for the elderly, susceptibility to falls and resulting injuries is no laughing matter.
Now, a development effort at Texas Tech University, sponsored by Texas Instruments Inc., is taking aim at preventing falls by analyzing posture and gait to send warning alerts to the elderly before they fall. The project has already enlisted volunteers at Texas Tech's Health Science Center to perfect the wireless wearable sensor and is on track next for clinical trials at the first U.S. on-campus geriatric teaching facility, Texas Tech's Geriatric Education and Care Center.
"We have tried all sorts of sensor placements, from leg-mounted inertial sensors to slippers with pressure sensors in their soles, but Texas Instruments' low-power microcontrollers and wireless radio combined with a torso-mounted MEMS accelerometer and gyro—like an external pacemaker—gives the best results," said lead scientist on the project, Texas Tech EE professor Donald Lie.
After a three-year development effort, Lie's team has crafted not only the torso-mounted wireless sensor but also the software analytics running on a PC that wirelessly monitors patients, resulting in algorithms that can reliably detect falls regardless of in which direction, which Lie claims is difficult to differentiate from the many ways that people plop down into their beds, cars seats and the various pieces of furniture around the home.
Prototype of wireless sensor module for detecting falls in geriatric patients is small enough to clip on a belt.
Photo courtesy of Texas Tech
Next, the team is honing its software analytics even finer to detect pre-fall conditions in order to take preventative measures by virtue of clinical trials on real patients at the local Geriatric Education and Care Center where they have already installed the necessary wireless infrastructure.
The MEMS sensor and wireless radio module is clipped on the belt, or for women not wearing a belt on the back of their bra, and contains a MEMS accelerometer, MEMS gyroscope, TI's MSP430 microcontroller and CC2500 radio frequency (RF) transceiver and uses the ultra-low-power SimpliciTI network protocol running a proprietary software stack developed on TI's ez430-RF2500 development platform.
"We became interested in this project at Texas Tech, because it went beyond just fall detection, but aspired to preventative measures," said TI Fellow Allen Bowling. "By analyzing the dynamics of posture and gait--the way people are standing or walking before a fall—we hope to be able to send an alarm instructing them to grab hold of something or sit down before they fall."
If the patient does end up falling, an alarm is sent wirelessly to a care provider, but the goal is to code the analytics—which today runs on a clinician-monitored PC—for the MSP430 microcontroller so alarms can be issued to patients regardless of whether they are in range of a wireless router or not.
The team also has its sights on monitoring other vestibular (balance related) disorders and diseases including Parkinson's, dementia and epilepsy, in hopes of issuing warnings to those patients of impending episodes thereby enabling them to take preventative measures.
"We believe that these projects will really make a significant difference in the clinical care of geriatric populations," said Lie, who credits his team of collaborators, including doctors Tam Nguyen, Steven Zupancic, Andrew Dentino, Ron Banister, and Tim Dallas.