Visualisation of Movement of Older Adults within their Homes based on PIR Sensor Data

SummaryBackground: Many older adults in the US prefer to live independently for as long as they are able, despite the onset of conditions such as frailty and dementia. Solutions are needed to enable independent living, while enhancing safety and peace of mind for their families. Elderly patients are particularly at-risk for late assessment of cognitive changes.Objectives: We predict early signs of illness in older adults by using the data generated by a continuous, unobtrusive nursing home monitoring system.Methods: We describe the possibility of employing a multiple instance learning (MIL) framework for early illness detection. The MIL framework is suitable for training classifiers when the available data presents temporal or location uncertainties.Results: We provide experiments on three datasets that prove the utility of the MIL framework. We first tuned our algorithms on a set of 200 normal/abnormal behavior patterns produced by a dedicated simulator. We then conducted two retrospective studies on residents from the Tiger Place aging in place facility, aged over 70, which have been monitored with motion and bed sensors for over two years. The presence or absence of the illness was manually assessed based on the nursing visit reports.Conclusions: The use of simulated sensor data proved to be very useful for algorithm development and testing. The results obtained using MIL for six Tiger Place residents, an average area under the receiver operator characteristic curve (AROC) of 0.7, are promising. However, more sophisticated MIL classifiers are needed to improve the performance.

Download Full-text

POOR SLEEP QUALITY IS RELATED TO DECREASED WHITE MATTER INTEGRITY IN BRAIN NOCICEPTIVE PATHWAYS IN OLDER ADULTS

Innovation in Aging ◽

10.1093/geroni/igz038.1343 ◽

2019 ◽

Vol 3 (Supplement_1) ◽

pp. S367-S368

Author(s):

Anna R Egbert ◽

Ryan S Falck ◽

John R Best ◽

Linda Li ◽

Lynne Feehan ◽

...

Keyword(s):

Older Adults ◽

White Matter ◽

Sleep Quality ◽

Sleep Efficiency ◽

Sensor Data ◽

White Matter Integrity ◽

Poor Sleep ◽

Accelerometer Data ◽

Poor Sleep Quality ◽

Brain White Matter

Abstract Poor sleep quality, decreased physical activity (PA) and increased sedentary behavior (SB) are common characteristics of older adults. Notably, these factors play an important role in brain health. We examined the relationship between sleep quality, PA, SB and brain white matter integrity (WM) in older adults with osteoarthritis (OA). We retained data on 16 participants (mean age 60, SD=7.7) from a larger Monitor-OA cohort recruited from Metro Vancouver, BC, Canada. Sleep efficiency and duration, amount of time spent on PA and SB daily over a period of one week was acquired with an objective measure – the multi-sensor monitor SenseWear Mini which integrates tri-axial accelerometer data, physiological sensor data and personal demographic information. Brain WM tractography was calculated from fractional anisotropy data obtained with diffusion weighted magnetic resonance imaging. Voxelwise group-level statistics examined the effects of our variables of interest on the integrity of brain WM tracts while controlling for participants age. We found that lower sleep efficiency was related to decreased integrity in WM tracts of frontal, temporal lobes, precuneus and thalamus (Bonferroni corrected p<0.05). Shorter sleep was related to lower WM integrity in frontal regions, posterior cingulate and insula radiations (Bonferroni corrected p<0.05). No significant effects were noted for PA or SB. The identified brain regions are involved in sleep processes but further overlap with the nociceptive brain network. Our findings suggest that neural mechanisms related to sleep disturbance may also involve pain-related processing in older adults.

Download Full-text

Comparison of Embedded Sensor Data for Long-Term Care Residents Before and After Onset of the COVID-19 Pandemic

Innovation in Aging ◽

10.1093/geroni/igab046.360 ◽

2021 ◽

Vol 5 (Supplement_1) ◽

pp. 95-96

Author(s):

Erin Robinson ◽

Wenlong Wu ◽

Geunhye Park ◽

Gashaye M Tefera ◽

Kari Lane ◽

...

Keyword(s):

Mental Health ◽

Older Adults ◽

Long Term Care ◽

Sensor Data ◽

Term Care ◽

Health Concerns ◽

Front Door ◽

Depth Sensors ◽

Mental Health Concerns

Abstract Older adults have experienced greater isolation and mental health concerns during the COVID-19 pandemic. In long-term care (LTC) settings, residents have been particularly impacted due to strict lockdown policies. Little is known about how these policies have impacted older adults. This study leveraged existing research with embedded sensors installed in LTC settings, and analyzed sensor data of residents (N=30) two months pre/post the onset of the U.S. COVID-19 pandemic (1/13/20 to 3/13/20, 03/14/20 to 5/13/20). Data from three sensors (bed sensors, depth sensors, and motion sensors) were analyzed for each resident using paired t-tests, which generated information on the resident’s pulse, respiration, sleep, gait, and motion in entering/exiting their front door, living rooms, bedrooms, and bathrooms. A 14.4% decrease was observed in front door motion in the two months post-onset of the pandemic, as well as a 2.4% increase in average nighttime respiration, and a 7.6% increase in nighttime bed restlessness. Over half of our sample (68%) had significant differences (p<0.05) in restlessness. These results highlight the potential impact of the COVID-19 pandemic and social distancing policies on older adults living in LTC. While it is not surprising that significant differences were found in the front door motion sensor, the bed sensor data can potentially shed light on how sleep was impacted during this time. As older adults experienced additional mental health concerns during this time, their normal sleep patterns could have been affected. Implications could help inform LTC staff, healthcare providers, and self-management of health approaches among older adults.

Download Full-text

Automatic Quantification of Tandem Walking Using a Wearable Device: New Insights Into Dynamic Balance and Mobility in Older Adults

The Journals of Gerontology Series A ◽

10.1093/gerona/glaa235 ◽

2020 ◽

Vol 76 (1) ◽

pp. 101-107

Author(s):

Natalie Ganz ◽

Eran Gazit ◽

Nir Giladi ◽

Robert J Dawe ◽

Anat Mirelman ◽

...

Keyword(s):

Older Adults ◽

Postural Control ◽

Dynamic Balance ◽

Principal Component ◽

Adverse Outcomes ◽

Community Dwelling ◽

Sensor Data ◽

Quiet Standing ◽

Mobility Disability ◽

Community Dwelling Older Adults

Abstract Background Wearable sensors are increasingly employed to quantify diverse aspects of mobility. We developed novel tandem walking (TW) metrics, validated these measures using data from community-dwelling older adults, and evaluated their association with mobility disability and measures of gait and postural control. Methods Six hundred ninety-three community-dwelling older adults (age: 78.69 ± 7.12 years) wore a 3D accelerometer on their lower back while performing 3 tasks: TW, usual-walking, and quiet standing. Six new measures of TW were extracted from the sensor data along with the clinician’s conventional assessment of TW missteps (ie, trip other loss of balance in which recovery occurred to prevent a fall) and duration. Principal component analysis transformed the 6 new TW measures into 2 summary TW composite factors. Logistic regression models evaluated whether these TW factors were independently associated with mobility disability. Results Both TW factors were moderately related to the TW conventional measures (r < 0.454, p < .001) and were mildly correlated with usual-walking (r < 0.195, p < .001) and standing, postural control (r < 0.119, p < .001). The TW frequency composite factor (p = .008), but not TW complexity composite factor (p = .246), was independently associated with mobility disability in a model controlling for age, sex, body mass index, race, conventional measures of TW, and other measures of gait and postural control. Conclusions Sensor-derived TW metrics expand the characterization of gait and postural control and suggest that they reflect a relatively independent domain of mobility. Further work is needed to determine if these metrics improve risk stratification for other adverse outcomes (eg, falls and incident disability) in older adults.

Download Full-text

How Are You Doing? Enabling Older Adults to Enrich Sensor Data with Subjective Input

Human Behavior Understanding - Lecture Notes in Computer Science ◽

10.1007/978-3-319-24195-1_4 ◽

2015 ◽

pp. 39-51 ◽

Cited By ~ 2

Author(s):

Marije Kanis ◽

Saskia Robben ◽

Ben Kröse

Keyword(s):

Older Adults ◽

Sensor Data ◽

Subjective Input

Download Full-text

Predictors of real-life mobility in community-dwelling older adults: an exploration based on a comprehensive framework for analyzing mobility

European Review of Aging and Physical Activity ◽

10.1186/s11556-019-0225-2 ◽

2019 ◽

Vol 16 (1) ◽

Cited By ~ 3

Author(s):

Eleftheria Giannouli ◽

Michelle Pasquale Fillekes ◽

Sabato Mellone ◽

Robert Weibel ◽

Otmar Bock ◽

...

Keyword(s):

Physical Activity ◽

Older Adults ◽

Physical Functioning ◽

Real Life ◽

Cardiovascular Fitness ◽

Community Dwelling ◽

Sensor Data ◽

Maximum Temperature ◽

Life Space ◽

Community Dwelling Older Adults

Abstract Background Reduced mobility is associated with a plethora of adverse outcomes. To support older adults in maintaining their independence, it first is important to have deeper knowledge of factors that impact on their mobility. Based on a framework that encompasses demographical, environmental, physical, cognitive, psychological and social domains, this study explores predictors of different aspects of real-life mobility in community-dwelling older adults. Methods Data were obtained in two study waves with a total sample of n = 154. Real-life mobility (physical activity-based mobility and life-space mobility) was assessed over one week using smartphones. Active and gait time and number of steps were calculated from inertial sensor data, and life-space area, total distance, and action range were calculated from GPS data. Demographic measures included age, gender and education. Physical functioning was assessed based on measures of cardiovascular fitness, leg and handgrip strength, balance and gait function; cognitive functioning was assessed based on measures of attention and executive function. Psychological and social assessments included measures of self-efficacy, depression, rigidity, arousal, and loneliness, sociableness, perceived help availability, perceived ageism and social networks. Maximum temperature was used to assess weather conditions on monitoring days. Results Multiple regression analyses indicated just physical and psychological measures accounted for significant but rather low proportions of variance (5–30%) in real-life mobility. Strength measures were retained in most of the regression models. Cognitive and social measures did not remain as significant predictors in any of the models. Conclusions In older adults without mobility limitations, real-life mobility was associated primarily with measures of physical functioning. Psychological functioning also seemed to play a role for real-life mobility, though the associations were more pronounced for physical activity-based mobility than life-space mobility. Further factors should be assessed in order to achieve more conclusive results about predictors of real-life mobility in community-dwelling older adults.

Download Full-text

Using Big Data for Clinical Decision Making

Innovation in Aging ◽

10.1093/geroni/igaa057.563 ◽

2020 ◽

Vol 4 (Supplement_1) ◽

pp. 174-174

Author(s):

Diana Woods ◽

Maria Yefimova

Keyword(s):

Older Adults ◽

Decision Making ◽

Chronic Conditions ◽

Clinical Decision Making ◽

Clinical Decision ◽

Sensor Data ◽

Innovative Technology ◽

Multiple Chronic Conditions ◽

Daily Routine ◽

Frail Older Adults

Abstract The current workforce is ill prepared for the rise in Americans 65 and older from 46.3 million in 2010 to 98.2 million by 2050, a national increase of 112.2 % accompanied by increasing chronic conditions. The increase in older Americans, the prevalence of those with dementia, accompanied by behavioral symptoms of dementia (BSD) is increasing. Innovative technology may alert health providers to early signs of decline in frail older adults with multiple chronic conditions. Remote monitoring in the home and community living spaces can address complex care needs for older adults. Monitoring may identify and predict deviations in a person’s daily routine that herald a change in a chronic condition. We present two examples that can potentially assist in clinical decision making. The first exemplar used 24/7 sensor data to identify changes, potentially clinically significant, such that early intervention may prevent hospitalizations; the second exemplar presents the use of pattern recognition software (THEME TM) for temporal pattern analysis, to identify and quantify behavior patterns with regard to intensity, frequency and complexity, such that interventions may be individually tailored and timed. Clinical researchers and technology developers need to collaborate early in the process to consider the sources and frequency of clinical measures for meaningful predictions. One major challenge lies in the interpretation of the vast amounts of within individual data. Our insights strive to improve future interdisciplinary development of monitoring systems to support aging in place and support clinical decisions for timely and effective care for frail older adults.

Download Full-text