A study in the Journal of Clinical Psychiatry.
You are here
Let's craft a vision for computer vision -- applied to aging
Seeing you with smart eyes that don't blink. Maybe you saw the intriguing article about the numerous and varied applications for computer vision -- some in the market, some just a gleam in the research and vendor eye. But we all want to think about the uses of computer vision -- not just a web camera for Skyping with the grandkids or used for playing cool games, but rather, a camera integrated with specific software that can react to the images seen -- and help with task or make the environment safer.
The article mentioned the ability to detect distress, falls, or other disturbances in hospital rooms, reminding a care provider to wash her hands before leaving a room, and detecting changes in vital signs through face recognition. These are useful technologies for rehabs, nursing homes, assisted living, or wherever infection, sleep disturbance, and fall risk are such vexing problems. Here are a few more not-so-original ideas -- please augment and add:
1) Facial recognition -- personalized interactions. Wander-prevention technologies (ankle and wrist bracelets and the like) originated in prisons to track prisoners. Today, the 'Wander Guard' model is used in assisted living and nursing homes to alert staff when a resident with dementia heads out the door. But rather than just an alert, buzzing alarm, or door clicking locked, what if the exiting Mrs. Smith is also warmly encouraged to re-enter for a snack or her favorite movie? What if Mrs. Smith sits down in a chair in a common room, ready to watch TV -- she is the only person in the room, so the system automatically adjusts the content to her preferences with a displayed message -- "Returning to Animal Planet -- okay?"
2) More vital signs -- any fluids today? The more immobile, the greater the risk of dehydration and deterioration -- the GE 'smart patient room' research offers up multiple possibilities. Has Mrs. Jones been offered a sip of water today from the pitcher on her tray table? If Mrs. Jones, a resident at the end of a long nursing home corridor, can't reach the call button, then slips down in her bed or falls between bed and chair, is she safer with computer vision software that alerts an aide to come check on her outside of regular tasks and rounds?
3) Training new staff about you and your needs. One of the perpetual dilemmas in home care, assisted living and nursing homes is job turnover. It's an industry that has one of the lowest pay rates for those delivering the highest level of care to elderly, the home-bound and those with dementia. What if it was feasible to train new or temp staff about Mrs. Smith and Mrs. Jones -- through a quick lookup system (on a portable, a smart phone or a kiosk) accessed by temps, new hires, weekend and holiday staff? With the highest level of security as a default, the system characterizes Mrs. Smith or Mrs. Jones with a visual identity collected at group events like sitting and chatting on a porch, attending meals or participating in group activities. Mrs. Smith is the one who likes to dance; Mrs. Jones' favorite meal is chocolate cake; both adore knitting and doing puzzles, all collected through software that sees each resident and assigns and retains an easily accessed visual profile.
More ideas welcome!