Study finds AI keeps residents safe

A study highlights the importance of artificial intelligence to detect falls for safer practice and harm reduction among residents.

A study highlights the importance of artificial intelligence to detect falls for safer practice and harm reduction among residents.

The University of Melbourne pilot study, at a 170-bed aged care facility with 200 staff located in Melbourne, aimed to investigate the feasibility of non-contact smart sensor systems to monitor behaviour and detect falls among residents.

Two non-contact artificial intelligence smart-sensors were wall mounted into resident bedrooms, one which could detect the resident in bed and the other in the bathroom.

The research, published in the journal Health Technology and Informatics earlier this year, found that the use of smart sensors for falls monitoring offers increased safety for residents. However it noted successful implementation couldn’t be achieved without:

  • patient and family members’ confidence
  • adequate infrastructure and resourcing, such as connectivity, bandwidth and on-site expertise of the sensor system at the facility
  • positive involvement of facility staff on both front line and senior levels

Project partners of the research, University of Melbourne Health and Biomedical Informatics Centre senior researcher Dr Ann Borda; Austin Health and University of Melbourne director of Physiotherapy research Dr Cathy Said; University of Melbourne health librarian and research assistant Cecily Gilbert; University of Melbourne lecturer and honorary research fellow Frank Smolenaers; and Semantrix chief technology officer Michael McGrath said that the smart sensors could change how falls is detected in aged care.

“If validated, forms of smart technology could enable speedier falls detection, particularly with the possibility of capturing the exact sequence of events preceding a fall, slip or trip that could support falls prediction,” the researchers told Australian Ageing Agenda.

Using AI to understand human movement

The researchers said the smart sensors utilise on-board artificial intelligence processing to understand human movements.

“They are calibrated to identify activity-monitoring through detection of inferred movements and articulated movement patterns representing falls and other behaviours of interest or concern,” the researchers said.

The smart sensors meant aged care residents didn’t need to use wearable monitoring devices, and their falls could be better understood by the use of multiple cooperating sensors.

“Patients do not need to wear alert pendants or tags or carry a mobile device or wear a body-attached sensor device,” the team said.

“(Residents are) observed by not just one, but several cooperating sensors which understand what a fall and falls behaviour events look like.”

The sensors can also send alerts to nursing staff on duty when falls are detected.

The researchers said, “given the long-term negatives of long-lies of residential patients from undetected falls, the ability to immediately alert nursing staff to respond to falls and dangerous falls behaviours is a huge benefit to staff and aged residents, and in general falls prevention efforts.”

Further research and longer-term studies will be conducted on the implementation of non-contact smart sensors to detect falls in aged care facilities.

Access the paper, Non-contact sensor-based falls detection in residential aged care facilities: developing a real-life picture here.

Comment below to have your say on this story

Send us your news and tip-offs to 

Subscribe to Australian Ageing Agenda magazine and sign up to the AAA newsletter

Tags: Cecily Gilbert, Dr Ann Borda, Dr Cathy Said, Frank Smolenaers, Michael McGrath, Semantrix, university-of-melbourne,

Leave a Reply

Your email address will not be published. Required fields are marked *