Instrumentation & Measurement Magazine 24-6 - 75

replicated with research groups in other countries collaborating
with the CART initiative. Identifiable PHI is stored locally
in agreement with local privacy laws. Participants are coded
with randomized numbers and anonymized data can then be
associated with each participant and stored in either central or
local databases that can then be shared between research sites,
with appropriate permissions. Data access is controlled for individual
researchers to access their relevant study data, while
this standardized data collection and management system improves
the ability to share data collected at different sites with
a consistent methodology. To achieve this, the protocols for ensuring
data privacy and security for research participants have
been reviewed and approved by research ethic boards (REB)
at each collaborating institution. Although additional collaborations
with the CART initiative will require REB approval for
each individual project, the existing project protocols that meet
national and local regulations within other countries facilitate
the process for other researchers.
International Collaboration
The CART platform is currently being used in a number of
research settings and projects in North America, Australia
and Europe. The platform is employed in studies in Ottawa,
ON, Canada led by the AGE-WELL national innovation hub
Sensors and Analytics for Monitoring Mobility and Memory
(SAM3
, http://agewell-nih-sam3.ca/) and in partnership
with the University of Victoria, BC, Canada. The platform
provides a basis to test new Canadian smart home technologies
with real world participants to help validate the systems'
ability to support older adults to maintain independence and
age in place. The standardized data management and collection
backend facilitates data sharing between partners and
increases the potential sample size for analysis of sensor data.
A major goal of the project is the development of real-time
data analytics that provide the ability to evaluate and develop
smart home care solutions. Initial work is evaluating methods
in which the platform could provide summary and feedback
information to individuals with cognitive impairment and
their caregivers [37]. This builds on prior work done by SAM3
researchers in the development of responsive feedback using
smart home technology, such as a system to address nighttime
wandering behavior in individuals with dementia [38].
Novel analytic techniques and comparisons between
sensors are being explored using the high-frequency data collected
by the platform. Visualization techniques for big data
from the PIR sensors in Ottawa-based participants demonstrate
methods to view and detect changes in activity levels
after social isolation protocols were put in place due to COVID-19
[39]. The development of fusion based algorithms that
combine data from multiple types of sensors may provide
more accurate assessment of certain health outcomes, for example,
bedmat-based sensors used in the CART platform and
smart home systems demonstrate variable capabilities and
limitations [40] for the detection of sleep-related behaviors.
The CART platform is also deployed in Victoria, to assess
day-to-day cognitive and physical health in older adults
September 2021
with multimorbidity and/or MCI to inform large-scale implementations
to maintain the health and independence of older
adults. The Victoria team, based at the Institute on Aging and
Lifelong Health of the University of Victoria, has also developed
an iOS/Android mobile health platform, MyCogHealth,
to provide more sensitive and reliable assessments of cognitive
functioning and within-person change. The Victoria-based
CART platform is permitting the evaluation of mobile health
technologies using established parameters derived from the
CART platform. A major aim is the co-validation of within-person
dynamics, including multivariate change-points, across
an array of outcomes, risk, and contextual factors to identify
and potentially prevent or delay cognitive health and functional
changes.
Collaborations at the University of Minneapolis, MN, USA
and the Minneapolis Veterans Administration are evaluating
the ability of the platform to assess the ability of in-home and
mobile technologies to detect early cognitive and functional
changes in military veterans [41], a group at higher risk of cognitive
impairment and dementia.
Application of the CART platform for assessment of unique
features of the clinically heterogeneous frontotemporal dementias
(FTD) is being conducted using the CART platform
at the University of California, San Francisco, USA as a pilot
study of the larger national ALLFTD US study (https://
www.allftd.org/). In this study deployment of the platform to
patients with FTD, to specifically examine characteristic movement,
language, and cognitive changes that commonly emerge
as these syndromes evolve, will be the focus of this longitudinal
assessment.
A collaboration between CSIRO (Royal Brisbane and Women's
Hospital, Australia) and the Oregon Center for Aging &
Technology (ORCATECH) is focused on cross-platform comparability
and standardization of aging in place digital metrics
by collocating the CART platform with the CSIRO Smarter
Safer Homes platform [42] in respective homes in Brisbane,
Australia, and Portland, Oregon, USA. An important goal of
this collaboration is to observe how platforms perform in culturally
different locales, as well as how data derived from these
platforms that collect complementary activity and functional
data compare.
In the Toulouse region of France, the population-based INSPIRE
study (https://inspire.chu-toulouse.fr/en/about/) of
aging biomarkers associated with independent and healthy
aging has launched a substudy focused on digital functional
biomarkers that will be derived from the CART platform deployed
to the homes of INSPIRE participants. The digital
biomarkers will be correlated with extensive fluid and other
conventional biomarkers and clinical assessments collected
by the INSPIRE team, providing the opportunity to model
healthy aging informed by objective, real-world continuously
acquired functional outcomes.
Conclusions
Multiple challenges remain to ensure that data collected
from home-based platforms meets certain standards for data
IEEE Instrumentation & Measurement Magazine
75
https://www.allftd.org/ https://www.allftd.org/ http://agewell-nih-sam3.ca/ https://inspire.chu-toulouse.fr/en/about/
Instrumentation & Measurement Magazine 24-6

Table of Contents for the Digital Edition of Instrumentation & Measurement Magazine 24-6
