Inertial measurements of free-living activities: assessing mobility to predict falls.

An exploratory analysis was conducted into how simple features, from acceleration at the lower back and ankle during simulated free-living walking, stair ascent and descent, correlate with age, the overall fall risk from a clinically validated Physiological Profile Assessment (PPA), and its sub-components. Inertial data were captured from 92 older adults aged 78-95 (42 female, mean age 84.1, standard deviation 3.9 years). The dominant frequency, peak width from Welch's power spectral density estimate, and signal variance along each axis, from each sensor location and for each activity were calculated. Several correlations were found between these features and the physiological risk factors. The strongest correlations were from the dominant frequency at the ankle along the mediolateral direction during stair ascent (Spearman's correlation coefficient p = - 0.45) with anterioposterior sway, and signal variance of the anterioposterior acceleration at the lower back during stair descent (p = - 0.45) with age. These findings should aid future attempts to classify activities and predict falls in older adults, based on true free-living data from a range of activities.

A low-power fall detection algorithm based on triaxial acceleration and barometric pressure.

This paper proposes a low-power fall detection algorithm based on triaxial accelerometry and barometric pressure signals. The algorithm dynamically adjusts the sampling rate of an accelerometer and manages data transmission between sensors and a controller to reduce power consumption. The results of simulation show that the sensitivity and specificity of the proposed fall detection algorithm are both above 96% when applied to a previously collected dataset comprising 20 young actors performing a combination of simulated falls and activities of daily living. This level of performance can be achieved despite a 10.9% reduction in power consumption.

Pilot evaluation of an unobtrusive system to detect falls at nighttime.

Research shows that older people (aged 65 years and over) suffer many unintentional indoor falls which often lead to severe injuries. As a result of an increasingly aged population in developed countries, a sizable portion of healthcare funding is consumed in the treatment of fall-related injuries and associated long-term care. Detecting falls soon after they occur can be potentially live saving. In addition, early treatment of fall-related injuries can reduce treatment costs by minimizing health deterioration resulting from long periods spent incapacitated on the floor after a fall (a scenario known as a `long lie') and decreasing the number of hospital bed-days required. In this study, a previously proposed unobtrusive nighttime fall detection system based on wireless passive infrared sensors and furniture load sensors is evaluated in a pilot study involving three older subjects, monitored for a combined total of 174 days. No falls occurred during the study. The system reported a false alarm rate of 0.53 falls per day, which is comparable with similar unobtrusive and wearable sensor fall detection solutions.

Validation of an accelerometer-based fall prediction model.

Falls are a common and serious problem faced by older populations. There is a growing interest in estimating the risk of falling for older people using body-worn sensors and simple movement tasks, allowing appropriate fall prevention programs to be administered in a timely manner to the high risk population. This study investigated the capability and validity of using a waist-mounted triaxial accelerometer (TA) and a directed routine (DR) that includes three movement tasks to discriminate between fallers and non-fallers and between multiple fallers and non-multiple fallers. Data were collected from 98 subjects who were stratified into two separate groups, one for model training and the other for model validation. Logistic regression models were constructed using the TA features from the entire DR and from each single DR task, and were validated using unseen data. The best models were obtained using features from the alternate step test to classify between fallers and non-fallers with κ = 0.34-0.41, sensitivity = 68%-71% and specificity = 63%-73%. However, the overall validation performances were poor. The study emphasizes the importance of independent validation in fall prediction studies.

Design of an unobtrusive system for fall detection in multiple occupancy residences.

A small trial was conducted to examine the feasibility of detecting falls using a combination of ambient passive infrared (PIR) and pressure mat (PM) sensors in a home with multiple occupants. The key tracking method made use of graph theoretical concepts to track each individual in the residence and to monitor them independently for falls. The proposed algorithm attempts to recognize falls where the subject experiences a hard fall on an indoor surface that leads to loss of consciousness or an inability to get up from the floor without assistance, due to severe injuries. The sensitivity, specificity and accuracy of the algorithm in detecting falls are 85.00%, 80.00% and 82.86%, respectively.

Energy expenditure estimation during normal ambulation using triaxial accelerometry and barometric pressure.

Energy expenditure (EE) is an important parameter in the assessment of physical activity. Most reliable techniques for EE estimation are too impractical for deployment in unsupervised free-living environments; those which do prove practical for unsupervised use often poorly estimate EE when the subject is working to change their altitude by walking up or down stairs or inclines. This study evaluates the augmentation of a standard triaxial accelerometry waist-worn wearable sensor with a barometric pressure sensor (as a surrogate measure for altitude) to improve EE estimates, particularly when the subject is ascending or descending stairs. Using a number of features extracted from the accelerometry and barometric pressure signals, a state space model is trained for EE estimation. An activity classification algorithm is also presented, and this activity classification output is also investigated as a model input parameter when estimating EE. This EE estimation model is compared against a similar model which solely utilizes accelerometry-derived features. A protocol (comprising lying, sitting, standing, walking, walking up stairs, walking down stairs and transitioning between activities) was performed by 13 healthy volunteers (8 males and 5 females; age: 23.8 ± 3.7 years; weight: 70.5 ± 14.9 kg), whose instantaneous oxygen uptake was measured by means of an indirect calorimetry system (K4b(2), COSMED, Italy). Activity classification improves from 81.65% to 90.91% when including barometric pressure information; when analyzing walking activities alone the accuracy increases from 70.23% to 98.54%. Using features derived from both accelerometry and barometry signals, combined with features relating to the activity classification in a state space model, resulted in a VO(2) estimation bias of -0.00 095 and precision (1.96SD) of 3.54 ml min(-1) kg(-1). Using only accelerometry features gives a relatively worse performance, with a bias of -0.09 and precision (1.96SD) of 5.99 ml min(-1) kg(-1), with the largest errors due to an underestimation of VO(2) when walking up stairs.

Spectral analysis of accelerometry signals from a directed-routine for falls-risk estimation.

Injurious falls are a prevalent and serious problem faced by a growing elderly population. Accurate assessment and long-term monitoring of falls-risk could prove useful in the prevention of falls, by identifying those at risk of falling early so targeted intervention may be prescribed. Previous studies have demonstrated the feasibility of using triaxial accelerometry to estimate the risk of a person falling in the near future, by characterizing their movement as they execute a restricted sequence of predefined movements in an unsupervised environment, termed a directed routine. This study presents an improvement on this previously published system, which relied explicitly on time-domain features extracted from the accelerometry signals. The proposed improvement incorporates features derived from spectral analysis of the same accelerometry signals; in particular the harmonic ratios between signal harmonics and the fundamental frequency component are used. Employing these additional frequency-domain features, in combination with the previously reported time-domain features, an increase in the observed correlation with the clinical gold-standard risk of falling, from = 0:81 to = 0:96, was achieved when using manually annotated event segmentation markers; using an automated algorithm to segment the signals gave corresponding results of = 0:73 and = 0:99, before and after the inclusion of spectral features. The strong correlation with falls-risk observed in this preliminary study further supports the feasibility of using an unsupervised assessment of falls-risk in the home environment.

Classification between non-multiple fallers and multiple fallers using a triaxial accelerometry-based system.

Falls are a prominent problem facing older adults and a common cause of hospitalized injuries. Accurate falls-risk assessment and classification of falls-risk levels will provide useful information for the prevention of future falls. This study presents a triaxial accelerometer (TA) based two-class classifier, which discriminates between multiple fallers and non-multiple fallers, using a directed-routine (DR) movement test. One-hundred-and-twenty-six features were extracted from the accelerometry signals, recorded during the DR tests using a waist mounted TA, from 68 subjects. A linear multiple regression model was employed to map a subset of these features to an estimate of the number of previous falls experienced in the preceding twelve months. A simple threshold is applied to this estimated number of falls to create a basic linear discriminant classifier to separate multiple from non-multiple fallers. The system attained an accuracy of 71% in classifying the exact number of falls experienced in the last 12 months and 97% in identifying multiple fallers.

Automatic segmentation of triaxial accelerometry signals for falls risk estimation.

Falls-related injuries in the elderly population represent one of the most significant contributors to rising health care expense in developed countries. In recent years, falls detection technologies have become more common. However, very few have adopted a preferable falls prevention strategy through unsupervised monitoring in the free-living environment. The basis of the monitoring described herein was a self-administered directed-routine (DR) comprising three separate tests measured by way of a waist-mounted triaxial accelerometer. Using features extracted from the manually segmented signals, a reasonable estimate of falls risk can be achieved. We describe here a series of algorithms for automatically segmenting these recordings, enabling the use of the DR assessment in the unsupervised and home environments. The accelerometry signals, from 68 subjects performing the DR, were manually annotated by an observer. Using the proposed signal segmentation routines, an good agreement was observed between the manually annotated markers and the automatically estimated values. However, a decrease in the correlation with falls risk to 0.73 was observed using the automatic segmentation, compared to 0.81 when using markers manually placed by an observer.

Barometric pressure and triaxial accelerometry-based falls event detection.

Falls and fall related injuries are a significant cause of morbidity, disability, and health care utilization, particularly among the age group of 65 years and over. The ability to detect falls events in an unsupervised manner would lead to improved prognoses for falls victims. Several wearable accelerometry and gyroscope-based falls detection devices have been described in the literature; however, they all suffer from unacceptable false positive rates. This paper investigates the augmentation of such systems with a barometric pressure sensor, as a surrogate measure of altitude, to assist in discriminating real fall events from normal activities of daily living. The acceleration and air pressure data are recorded using a wearable device attached to the subject's waist and analyzed offline. The study incorporates several protocols including simulated falls onto a mattress and simulated activities of daily living, in a cohort of 20 young healthy volunteers (12 male and 8 female; age: 23.7 ±3.0 years). A heuristically trained decision tree classifier is used to label suspected falls. The proposed system demonstrated considerable improvements in comparison to an existing accelerometry-based technique; showing an accuracy, sensitivity and specificity of 96.9%, 97.5%, and 96.5%, respectively, in the indoor environment, with no false positives generated during extended testing during activities of daily living. This is compared to 85.3%, 75%, and 91.5% for the same measures, respectively, when using accelerometry alone. The increased specificity of this system may enhance the usage of falls detectors among the elderly population.

Longitudinal falls-risk estimation using triaxial accelerometry.

Falls among the elderly population are a major cause of morbidity and injury-particularly among the over 65 years age group. Validated clinical tests and associated models, built upon assessment of functional ability, have been devised to estimate an individual's risk of falling in the near future. Those identified as at-risk of falling may be targeted for interventative treatment. The migration of these clinical models estimating falls risk to a surrogate technique, for use in the unsupervised environment, might broaden the reach of falls-risk screening beyond the clinical arena. This study details an approach that characterizes the movements of 68 elderly subjects performing a directed routine of unsupervised physical tasks. The movement characterization is achieved through the use of a triaxial accelerometer. A number of fall-related features, extracted from the accelerometry signals, combined with a linear least squares model, maps to a clinically validated measure of falls risk with a correlation of rho = 0.81 (p < 0.001).

Evaluation of functional deficits and falls risk in the elderly--methods for preventing falls.

Falls in the elderly have a profound impact on their quality of life through injury, increased fear of falling, reduced confidence to perform daily tasks and loss of independence. Falls come at a substantial economic cost. Tools to quantify falls risk and evaluate functional deficits allow interventions to be targeted to those at increased risk of falling and tailored to correct deficits with the aim of reducing falls rate and reducing ones risk of falling. We describe a system to evaluate falls risk and functional deficits in the elderly. The system is based on the evaluation of performance in a simple set of controlled movements known as the directed routine (DR). We present preliminary results of the DR in a cohort of 68 subjects using features extracted from the DR. Linear least-squares models were trained to estimate falls risk, knee-extension strength, proprioception, mediolateral body sway, anteroposterior body sway and contrast sensitivity. The model estimates provided good to fair correlations with (r=0.76 p<0.001), (r=0.65 p<0.001), (r=0.35 p<0.01), (r=0.53 p<0.001), (r=0.48 p<0.001) and (r=0.37 p<0.01) respectively.

Falls event detection using triaxial accelerometry and barometric pressure measurement.

A falls detection system, employing a Bluetooth-based wearable device, containing a triaxial accelerometer and a barometric pressure sensor, is described. The aim of this study is to evaluate the use of barometric pressure measurement, as a surrogate measure of altitude, to augment previously reported accelerometry-based falls detection algorithms. The accelerometry and barometric pressure signals obtained from the waist-mounted device are analyzed by a signal processing and classification algorithm to discriminate falls from activities of daily living. This falls detection algorithm has been compared to two existing algorithms which utilize accelerometry signals alone. A set of laboratory-based simulated falls, along with other tasks associated with activities of daily living (16 tests) were performed by 15 healthy volunteers (9 male and 6 female; age: 23.7 +/- 2.9 years; height: 1.74 +/- 0.11 m). The algorithm incorporating pressure information detected falls with the highest sensitivity (97.8%) and the highest specificity (96.7%).

A wearable triaxial accelerometry system for longitudinal assessment of falls risk.

Falls-related injuries in the elderly population are a major cause of morbidity and represent one of the most significant contributors to hospitalizations and rising health care expense in developed countries. Many laboratory-based studies have described falls detection systems using wearable accelerometry. However, only a limited number of reports have tried to address the difficult issues of falls detection and falls prevention in unsupervised or free-living environments. We describe a waist-mounted triaxial accelerometry (Triax) system with a remote data collection capability to provide unsupervised monitoring of the elderly. The basis of the monitoring is a self-administered directed-routine (DR) comprising three separate tests measured by way of the Triax. We present an initial evaluation of the DR results in 36 patients to detect early changes in functional ability and facilitate falls risk stratification. Extracted features considered alone show a correlation with falls risk of approximately rho=0.5. Estimation of falls risk using a linear least squares model provides a root-mean-squared error of 0.69 (rho=0.58, p<0.0002).

Falls management: detection and prevention, using a waist-mounted triaxial accelerometer.

We describe a distributed falls management system capable of real-time falls detection in an unsupervised living context and remote longitudinal tracking of falls risk parameters using a waist-mounted triaxial accelerometer. A self-administrable falls risk assessment is used to facilitate falls prevention. A web-interface allows clinicians to monitor the status of individuals and track their compliance with exercise interventions. Early identification of increased falls risk allows targeted interventions to be promptly administered. Real-time detection of falls allows immediate emergency response protocols to be deployed, reducing morbidity and increasing the independence of the community-dwelling elderly community.

Implementation of a real-time human movement classifier using a triaxial accelerometer for ambulatory monitoring.

The real-time monitoring of human movement can provide valuable information regarding an individual's degree of functional ability and general level of activity. This paper presents the implementation of a real-time classification system for the types of human movement associated with the data acquired from a single, waist-mounted triaxial accelerometer unit. The major advance proposed by the system is to perform the vast majority of signal processing onboard the wearable unit using embedded intelligence. In this way, the system distinguishes between periods of activity and rest, recognizes the postural orientation of the wearer, detects events such as walking and falls, and provides an estimation of metabolic energy expenditure. A laboratory-based trial involving six subjects was undertaken, with results indicating an overall accuracy of 90.8% across a series of 12 tasks (283 tests) involving a variety of movements related to normal daily activities. Distinction between activity and rest was performed without error; recognition of postural orientation was carried out with 94.1% accuracy, classification of walking was achieved with less certainty (83.3% accuracy), and detection of possible falls was made with 95.6% accuracy. Results demonstrate the feasibility of implementing an accelerometry-based, real-time movement classifier using embedded intelligence.