Grant: Canada Postdoctoral Research Award (CPRA) program
University of Toronto
$140,000.00
Personalized Fall Risk Assessment through In-Home Privacy-Respecting Sensing Technologies
FY 2025-26
Summary
Falls are one of the biggest threats to health and independence in older adults. A single fall can cause fractures, hospitalizations, and long-term loss of mobility. Yet most fall-risk assessments are done in clinics using brief walking or balance tests. These snapshots don't reflect everyday challenges at home-moving through cluttered rooms, navigating stairs, or walking when tired. To prevent falls, we need to measure risk where it matters: in the home.
There are two main ways to assess gait at home. Wearables (e.g., accelerometers or smartwatches) are easy to pilot but hard to sustain-people may forget to wear or charge them, placement varies, and they often miss joint-level motion details. Non-wearables like cameras or LiDAR can estimate detailed movement, but they have a limited field of view, so many units are required to cover an entire home. They also need clear lines of sight, are sensitive to lighting and clutter, and raise privacy concerns.
We will develop a third approach: a non-wearable, camera-free system that uses ultra-wideband (UWB) radio signals to monitor movement. UWB emits tiny, low-power pulses-similar in safety to everyday Wi-Fi-and measures how they reflect from the body. Unlike cameras, it captures no images or sound. Unlike wearables, there is nothing to remember or recharge. It can sense movement in darkness and through walls, so with only a few devices it can cover an entire home-making it especially practical for real-world use
From UWB, we will estimate 3-D body pose and predict who is at high risk of falling in the near future. We will evaluate accuracy, reliability across different households, and fairness across participant groups, and co-design simple installation and interpretation workflows with clinicians and caregivers.
Impact: A safe, affordable, privacy-protecting tool that identifies fall risk early, enables timely interventions (e.g., physiotherapy or medication review), and helps older adults live independently at home.
Department
Program
Canada Postdoctoral Research Award (CPRA) program
Recipient
Location
Ontario, Canada
Recipient Type
Academia
Type
Grant
Source
CIHR Funding Decisions Database
As published
Id
537721
Name
Dousty Mehdy
Region
Ontario
Country
Canada
Orgtype
Institution Paid; Research Institution
Instname2
Aging
Theme2
Biomedical
Prcname2
Canada Postdoctoral Research Award - A
Projecttitle
Personalized Fall Risk Assessment through In-Home Privacy-Respecting Sensing Technologies
Programtype2
Fellowship Programs
Orgnameinp2
University of Toronto
Cihramount2
$140,000
Cihrequipment2
$0
Programname2
Canada Postdoctoral Research Award (CPRA) program
Keyworddelim
Complex Neural Network; Computer Vision; Deep Learning; Fall Prevention; Falls; Gait Analysis; Machine Learning; Radar Sensing
Abstract
Falls are one of the biggest threats to health and independence in older adults. A single fall can cause fractures, hospitalizations, and long-term loss of mobility. Yet most fall-risk assessments are done in clinics using brief walking or balance tests. These snapshots don't reflect everyday challenges at home-moving through cluttered rooms, navigating stairs, or walking when tired. To prevent falls, we need to measure risk where it matters: in the home.
There are two main ways to assess gait at home. Wearables (e.g., accelerometers or smartwatches) are easy to pilot but hard to sustain-people may forget to wear or charge them, placement varies, and they often miss joint-level motion details. Non-wearables like cameras or LiDAR can estimate detailed movement, but they have a limited field of view, so many units are required to cover an entire home. They also need clear lines of sight, are sensitive to lighting and clutter, and raise privacy concerns.
We will develop a third approach: a non-wearable, camera-free system that uses ultra-wideband (UWB) radio signals to monitor movement. UWB emits tiny, low-power pulses-similar in safety to everyday Wi-Fi-and measures how they reflect from the body. Unlike cameras, it captures no images or sound. Unlike wearables, there is nothing to remember or recharge. It can sense movement in darkness and through walls, so with only a few devices it can cover an entire home-making it especially practical for real-world use
From UWB, we will estimate 3-D body pose and predict who is at high risk of falling in the near future. We will evaluate accuracy, reliability across different households, and fairness across participant groups, and co-design simple installation and interpretation workflows with clinicians and caregivers.
Impact: A safe, affordable, privacy-protecting tool that identifies fall risk early, enables timely interventions (e.g., physiotherapy or medication review), and helps older adults live independently at home.
Competitiondate
202509
Approvedterm2
2 yrs 0 mth
Pinamesdelim
Dousty, Mehdy
Approvedterm
2 ans 0 mois
Cihrcontribution2
~ $140,000~ $0~