New AI technology helps to reduce malnutrition in long-term care homes

New technology can automatically reduce and eliminate malnutrition and improve overall health by automatically documenting and monitoring how much food a resident consumes.

The smart system, developed by researchers at the University of Waterloo, the Schleigh-Jubilee Research Institute for Aging and researchers at the University Health Network, uses artificial intelligence software to analyze photographs of food after its residents have eaten.

Advanced software that analyzes color, depth, and other photo features can estimate how much each type of food is used and calculate the nutritional value.

Nowadays, it is not uncommon for a resident to eat only protein or carbohydrates.. “

Kaylen Pfisterer, Associate Research Fellow, University of Waterloo

He is currently pursuing a PhD in Water Design System Design Engineering.

“Our system is linked to long-term care at home and uses artificial intelligence to monitor how much food is being consumed to ensure its residents are meeting their unique needs.”

It is estimated that more than half the population in long-term care homes is at risk of malnutrition.

Food intake is now primarily controlled by staff who record consumption estimates by observing residents’ finished meals.

Robert Amylard, a student at Water University and a postdoctoral fellow at the University Health Network, says the subject matter of the process is 50 percent or more erroneous. In contrast, the automated system is five percent accurate, “providing good information on consumption patterns.”

Researchers collaborate with personal caregivers, nutritionists, and other long-term caregivers to develop the system, which saves time and improves accuracy and can be added to tablet computers used by frontline staff to capture electronic records.

“My vision is to monitor changes in the diet, such as the yellow or red flag, to monitor the health and infection control of its residents,” he said, now an associate professor of international science at the university’s health network. eHealth Innovation.


Journal reference dead

Pifster, KJ, Inter alia. (2022) Automated Nutrition Monitoring requires an in-depth translation section to address vision differences in long-term care settings. Scientific reports.


Leave a Comment