Clinical classification of memory and cognitive impairment with multimodal digital biomarkers

INTRODUCTION

Early detection of Alzheimer's disease and cognitive impairment is critical to improving the healthcare trajectories of aging adults, enabling early intervention and potential prevention of decline.

METHODS

To evaluate multi-modal feature sets for assessing memory and cognitive impairment, feature selection and subsequent logistic regressions were used to identify the most salient features in classifying Rey Auditory Verbal Learning Test-determined memory impairment.

RESULTS

Multimodal models incorporating graphomotor, memory, and speech and voice features provided the stronger classification performance (area under the curve = 0.83; sensitivity = 0.81, specificity = 0.80). Multimodal models were superior to all other single modality and demographics models.

DISCUSSION

The current research contributes to the prevailing multimodal profile of those with cognitive impairment, suggesting that it is associated with slower speech with a particular effect on the duration, frequency, and percentage of pauses compared to normal healthy speech.

Fall risk stratification of community-living older people. Commentary on the world guidelines for fall prevention and management

The Task Force on Global Guidelines for Falls in Older Adults has put forward a fall risk stratification tool for community-dwelling older adults. This tool takes the form of a flowchart and is based on expert opinion and evidence. It divides the population into three risk categories and recommends specific preventive interventions or treatments for each category. In this commentary, we share our insights on the design, validation, usability and potential impact of this fall risk stratification tool with the aim of guiding future research.

Estimating balance, cognitive function, and falls risk using wearable sensors and the sit-to-stand test

The five times sit-to-stand test (FTSS) is an established functional test, used clinically as a measure of lower-limb strength, endurance and falls risk. We report a novel method to estimate and classify cognitive function, balance impairment and falls risk using the FTSS and body-worn inertial sensors. 168 community dwelling older adults received a Comprehensive Geriatric Assessment which included the Mini-Mental State Examination (MMSE) and the Berg Balance Scale (BBS). Each participant performed an FTSS, with inertial sensors on the thigh and torso, either at home or in the clinical environment. Adaptive peak detection was used to identify phases of each FTSS from torso or thigh-mounted inertial sensors. Features were then extracted from each sensor to quantify the timing, postural sway and variability of each FTSS. The relationship between each feature and MMSE and BBS was examined using Spearman's correlation. Intraclass correlation coefficients were used to examine the intra-session reliability of each feature. A Poisson regression model with an elastic net model selection procedure was used to estimate MMSE and BBS scores, while logistic regression and sequential forward feature selection was used to classify participants according to falls risk, cognitive decline and balance impairment. BBS and MMSE were estimated using cross-validation with low root mean squared errors of 2.91 and 1.50, respectively, while the cross-validated classification accuracies for balance impairment, cognitive decline, and falls risk were 81.96, 72.71, and 68.74%, respectively. The novel methods reported provide surrogate measures which may have utility in remote assessment of physical and cognitive function.

Development of Data-driven Metrics for Balance Impairment and Fall Risk Assessment in Older Adults

Ageing incurs a natural decline of postural control which has been linked to an increased risk of falling. Accurate balance assessment is important in identifying postural instability and informing targeted interventions to prevent falls in older adults. Inertial sensor (IMU) technology offers a low-cost means for objective quantification of human movement. This paper describes two studies carried out to advance the use of IMU-based balance assessments in older adults. Study 1 (N=39) presents the development of two new IMU-derived balance measures. Study 2 (N=248) reports a reliability analysis of IMU postural stability measures and validates the novel balance measures through comparison with clinical scales. We also report a statistical fall risk estimation algorithm based on IMU data captured during static balance assessments alongside a method of improving this fall risk estimate by incorporating standard clinical fall risk factor data. Results suggest that both new balance measures are sensitive to balance deficits captured by the Berg Balance Scale (BBS) and Timed Up and Go test. Results obtained from the fall risk classifier models suggest they are more accurate (67.9%) at estimating fall risk status than a model based on BBS (59.2%). While the accuracies of the reported models are lower than others reported in the literature, the simplicity of the assessment makes it a potentially useful screening tool for balance impairments and falls risk. The algorithms presented in this paper may be suitable for implementation on a smartphone and could facilitate unsupervised assessment in the home.

Predicting Fall Counts Using Wearable Sensors: A Novel Digital Biomarker for Parkinson's Disease

People with Parkinson's disease (PD) experience significant impairments to gait and balance; as a result, the rate of falls in people with Parkinson's disease is much greater than that of the general population. Falls can have a catastrophic impact on quality of life, often resulting in serious injury and even death. The number (or rate) of falls is often used as a primary outcome in clinical trials on PD. However, falls data can be unreliable, expensive and time-consuming to collect. We sought to validate and test a novel digital biomarker for PD that uses wearable sensor data obtained during the Timed Up and Go (TUG) test to predict the number of falls that will be experienced by a person with PD. Three datasets, containing a total of 1057 (671 female) participants, including 71 previously diagnosed with PD, were included in the analysis. Two statistical approaches were considered in predicting falls counts: the first based on a previously reported falls risk assessment algorithm, and the second based on elastic net and ensemble regression models. A predictive model for falls counts in PD showed a mean R² value of 0.43, mean error of 0.42 and a mean correlation of 30% when the results were averaged across two independent sets of PD data. The results also suggest a strong association between falls counts and a previously reported inertial sensor-based falls risk estimate. In addition, significant associations were observed between falls counts and a number of individual gait and mobility parameters. Our preliminary research suggests that the falls counts predicted from the inertial sensor data obtained during a simple walking task have the potential to be developed as a novel digital biomarker for PD, and this deserves further validation in the targeted clinical population.

Unsupervised Assessment of Balance and Falls Risk Using a Smartphone and Machine Learning

Assessment of health and physical function using smartphones (mHealth) has enormous potential due to the ubiquity of smartphones and their potential to provide low cost, scalable access to care as well as frequent, objective measurements, outside of clinical environments. Validation of the algorithms and outcome measures used by mHealth apps is of paramount importance, as poorly validated apps have been found to be harmful to patients. Falls are a complex, common and costly problem in the older adult population. Deficits in balance and postural control are strongly associated with falls risk. Assessment of balance and falls risk using a validated smartphone app may lessen the need for clinical assessments which can be expensive, requiring non-portable equipment and specialist expertise. This study reports results for the real-world deployment of a smartphone app for self-directed, unsupervised assessment of balance and falls risk. The app relies on a previously validated algorithm for assessment of balance and falls risk; the outcome measures employed were trained prior to deployment on an independent data set. Results for a sample of 594 smartphone assessments from 147 unique phones show a strong association between self-reported falls history and the falls risk and balance impairment scores produced by the app, suggesting they may be clinically useful outcome measures. In addition, analysis of the quantitative balance features produced seems to suggest that unsupervised, self-directed assessment of balance in the home is feasible.

Reliability of inertial sensor based spatiotemporal gait parameters for short walking bouts in community dwelling older adults

BACKGROUND

When performing quantitative analysis of gait in older adults we need to strike a balance between capturing sufficient data for reliable measurement and avoiding issues such as fatigue. The optimal bout duration is that which contains sufficient gait cycles to enable a reliable and representative estimate of gait performance.

RESEARCH QUESTION

How does the number of gait cycles in a walking bout influence reliability of spatiotemporal gait parameters measured using body-worn inertial sensors in a cohort of community dwelling older adults?

METHODS

One hundred and fifteen (115) community dwelling older adults executed three 30-metre walk trials in a single measurement session. Bilateral gait data were collected using two inertial sensors attached to each participant's right and left shank, and gait events detected from the medio-lateral angular velocity signal. The number of gait cycles selected from each walking trial was varied from 3 to 16. Intraclass correlation coefficients (ICC(2,k)) were calculated to evaluate the reliability of each spatiotemporal gait parameter according to the number of gait cycles included in the analysis.

RESULTS

The specified algorithm and the clipping procedure for extracting short bouts of gait data seem appropriate for assessing older adults, providing reliable spatiotemporal measures from three gait cycles (three strides per leg) and good reliability for most parameters describing gait variability and gait asymmetry after six gait cycles (six strides per leg).

SIGNIFICANCE

A combination of using bilateral sensor data and adaptive thresholds for gait event detection enable reliable measures of spatiotemporal gait parameters over short walking bouts (minimum six gait cycles) in community dwelling older adults. This opens new possibilities in the use of wearable sensors in gait assessment based on short walking tasks. We recommend the number of gait cycles should be reported along with the calculated measures as reference values.

Short Bouts of Gait Data and Body-Worn Inertial Sensors Can Provide Reliable Measures of Spatiotemporal Gait Parameters from Bilateral Gait Data for Persons with Multiple Sclerosis

Wearable devices equipped with inertial sensors enable objective gait assessment for persons with multiple sclerosis (MS), with potential use in ambulatory care or home and community-based assessments. However, gait data collected in non-controlled settings are often fragmented and may not provide enough information for reliable measures. This paper evaluates a novel approach to (1) determine the effects of the length of the walking task on the reliability of calculated measures and (2) identify digital biomarkers for gait assessments from fragmented data. Thirty-seven participants (37) diagnosed with relapsing-remitting MS (EDSS range 0 to 4.5) executed two trials, walking 20 m each, with inertial sensors attached to their right and left shanks. Gait events were identified from the medio-lateral angular velocity, and short bouts of gait data were extracted from each trial, with lengths varying from 3 to 9 gait cycles. Intraclass correlation coefficients (ICCs) evaluate the degree of agreement between the two trials of each participant, according to the number of gait cycles included in the analysis. Results show that short bouts of gait data, including at least six gait cycles of bilateral data, can provide reliable gait measurements for persons with MS, opening new perspectives for gait assessment using fragmented data (e.g., wearable devices, community assessments). Stride time variability and asymmetry, as well as stride velocity variability and asymmetry, should be further explored as digital biomarkers to support the monitoring of symptoms of persons with neurological diseases.

Impact of Exercise Intervention in Parkinson's Disease can be Quantified Using Inertial Sensor Data and Clinical Tests

Parkinson's Disease (PD) has the second-highest prevalence rate of all neurodegenerative disorders. It effects approximately 1% of the population over the age of 60, with this proportion rising further, in more elderly cohorts. PD manifests as several motor and non-motor disfunctions, which develop progressively over time. Gait and mobility problems are amongst the most debilitating symptoms for people with PD. They severely affect a person's ability to carry out daily activities of living and can lead to a decreased quality of life. However, recent research has shown exercise intervention to be effective in improving gait, and overall functional mobility, in persons with PD. In this paper, we study the effect of an exercise intervention, comprised of three separate methods of exercise - all which have been shown previously to be effective individually - on a cohort with early-to-moderate stage PD. We also examine the ability of the Timed Up and Go (TUG) test - instrumented with inertial sensors (QTUG) - and the Unified Parkinson's Disease Rating Scale (UPDRS) Part III in measuring the response to the exercise intervention. We found that TUG time and the QTUG-derived frailty index - along with many additional parameters derived from QTUG - showed a significant change between baseline and post-intervention, while the UPDRS Part III score did not. The direction of the changes in the QTUG parameters also align with the expected exercise effect from the literature. Our results suggest QTUG may be a more sensitive measure than UPDRS Part III for assessing the effect of exercise intervention on functional mobility in people with early-to-moderate stage PD.

Investigating normal day to day variations of postural control in a healthy young population using Wii balance boards

The quantification of postural control (PC) provides the opportunity to understand the function and integration of the sensorimotor subsystems. The increased availability of portable sensing technology, such as Wii Balance Boards (WBB), has afforded the capacity to capture data pertaining to motor function, outside of the laboratory and clinical setting. However, prior to its use in long-term monitoring, it is crucial to understand natural daily PC variation. Twenty-four young adults conducted repeated static PC assessments over 20 consecutive weekdays, using WBBs. 16/24 participants (eyes open) and 11/24 participants (eyes closed) exhibited statistically significant differences (p <; 0.05) between their initial `once-off' measure and their daily measures of PC. This study showed that variations in PC exist in a healthy population, a once-off measure may not be representative of true performance and this inherent variation should be considered when implementing long-term monitoring protocols.

SENSOR-BASED ASSESSMENT OF FALLS RISK OF THE TIMED UP AND GO IN REAL-WORLD SETTINGS

Falls are the leading cause of older adult injury and cost $50bn annually. New digital technologies can quantitatively measure falls risk. Objective is to report on a validated wearable sensor-based Timed Up and Go (QTUG) assessment detailing 11 measures of falls risk, frailty and mobility impairment in older adults in six countries in 38 clinical and community settings. Second objective is to generate individual targeted falls prevention programs. 14,611 QTUG records from 8,521 participants (63% female) (72.7±10.7 years) available for analysis. QTUG time was 13.9±7.4 s; gait velocity was 101.9±32.5 cm/s. 25.8% of patients reported falling in previous 12 months; 26.2% of patients were at high fall risk. 21.5% not reporting a fall, were high fall risk. Participants had slow walking speed (29.8%); high gait variability (19.8%); problems with transfers (17.5%). Easily captured and interpreted sensor data is useful in a population-based approach to quantify falls risk stratification.

Longitudinal assessment of falls in patients with Parkinson's disease using inertial sensors and the Timed Up and Go test

Objective

To examine the predictive validity of a TUG test for falls risk, quantified using body-worn sensors (QTUG) in people with Parkinson's Disease (PD). We also sought to examine the inter-session reliability of QTUG sensor measures and their association with the Unified Parkinson's Disease Rating Scale (UPDRS) motor score.

Approach

A six-month longitudinal study of 15 patients with Parkinson's disease. Participants were asked to complete a weekly diary recording any falls activity for six months following baseline assessment. Participants were assessed monthly, using a Timed Up and Go test, quantified using body-worn sensors, placed on each leg below the knee.

Main results

The results suggest that the QTUG falls risk estimate recorded at baseline is 73.33% (44.90, 92.21) accurate in predicting falls within 90 days, while the Timed Up and Go time at baseline was 46.67% (21.27, 73.41) accurate. The Timed Up and Go time and QTUG falls risk estimate were strongly correlated with UPDRS motor score. Fifty-two of 59 inertial sensor parameters exhibited excellent inter-session reliability, five exhibited moderate reliability, while two parameters exhibited poor reliability.

Significance

The results suggest that QTUG is a reliable tool for the assessment of gait and mobility in Parkinson's disease and, furthermore, that it may have utility in predicting falls in patients with Parkinson's disease.

Fall Risk Assessment Through Automatic Combination of Clinical Fall Risk Factors and Body-Worn Sensor Data

Falls are the leading global cause of accidental death and disability in older adults and are the most common cause of injury and hospitalization. Accurate, early identification of patients at risk of falling, could lead to timely intervention and a reduction in the incidence of fall-related injury and associated costs. We report a statistical method for fall risk assessment using standard clinical fall risk factors (N = 748). We also report a means of improving this method by automatically combining it, with a fall risk assessment algorithm based on inertial sensor data and the timed-up-and-go test. Furthermore, we provide validation data on the sensor-based fall risk assessment method using a statistically independent dataset. Results obtained using cross-validation on a sample of 292 community dwelling older adults suggest that a combined clinical and sensor-based approach yields a classification accuracy of 76.0%, compared to either 73.6% for sensor-based assessment alone, or 68.8% for clinical risk factors alone. Increasing the cohort size by adding an additional 130 subjects from a separate recruitment wave (N = 422), and applying the same model building and validation method, resulted in a decrease in classification performance (68.5% for combined classifier, 66.8% for sensor data alone, and 58.5% for clinical data alone). This suggests that heterogeneity between cohorts may be a major challenge when attempting to develop fall risk assessment algorithms which generalize well. Independent validation of the sensor-based fall risk assessment algorithm on an independent cohort of 22 community dwelling older adults yielded a classification accuracy of 72.7%. Results suggest that the present method compares well to previously reported sensor-based fall risk assessment methods in assessing falls risk. Implementation of objective fall risk assessment methods on a large scale has the potential to improve quality of care and lead to a reduction in associated hospital costs, due to fewer admissions and reduced injuries due to falling.

Assessment and classification of early-stage multiple sclerosis with inertial sensors: comparison against clinical measures of disease state

A cross-sectional study on patients with early-stage multiple sclerosis (MS) was conducted to examine the reliability of manual and automatic mobility measures derived from shank-mounted inertial sensors during the Timed Up and Go (TUG) test, compared to control subjects. Furthermore, we aimed to determine if disease status [as measured by the Multiple Sclerosis Impact Scale (MSIS-20) and the Expanded Disability Status Score (EDSS)] can be explained by measurements obtained using inertial sensors. We also aimed to determine if patients with early-stage MS could be automatically distinguished from healthy controls subjects, using inertial parameters recorded during the TUG test. The mobility of 38 patients (aged 25-65 years, 14 M, 24 F), diagnosed with relapsing-remitting MS and 33 healthy controls (14 M, 19 F, age 50-65), was assessed using the TUG test, while patients wore inertial sensors on each shank. Reliability analysis showed that 36 of 53 mobility parameters obtained during the TUG showed excellent intrasession reliability, while nine of 53 showed moderate reliability. This compared favorably with the reliability of the mobility parameters in healthy controls. Exploratory regression models of the EDSS and MSIS-20 scales were derived, using mobility parameters and an elastic net procedure in order to determine which mobility parameters influence disease state. A cross-validated elastic net regularized regression model for MSIS-20 yielded a mean square error (MSE) of 1.1 with 10 degrees of freedom (DoF). Similarly, an elastic net regularized regression model for EDSS yielded a cross-validated MSE of 1.3 with 10 DoF. Classification results show that the mobility parameters of participants with early-stage MS could be distinguished from controls with 96.90% accuracy. Results suggest that mobility parameters derived from MS patients while completing the TUG test are reliable, are associated with disease state in MS, and may have utility in screening for early-stage MS.

Classification of frailty and falls history using a combination of sensor-based mobility assessments

Frailty is an important geriatric syndrome strongly linked to falls risk as well as increased mortality and morbidity. Taken alone, falls are the most common cause of injury and hospitalization and one of the principal causes of death and disability in older adults worldwide. Reliable determination of older adults' frailty state in concert with their falls risk could lead to targeted intervention and improved quality of care. We report a mobile assessment platform employing inertial and pressure sensors to quantify the balance and mobility of older adults using three physical assessments (timed up and go (TUG), five times sit to stand (FTSS) and quiet standing balance). This study examines the utility of each individual assessment, and the novel combination of assessments, to screen for frailty and falls risk in older adults.Data were acquired from inertial and pressure sensors during TUG, FTSS and balance assessments using a touchscreen mobile device, from 124 community dwelling older adults (mean age 75.9 ± 6.6 years, 91 female). Participants were given a comprehensive geriatric assessment which included questions on falls and frailty. Methods based on support vector machines (SVM) were developed using sensor-derived features from each physical assessment to classify patients at risk of falls risk and frailty.In classifying falls history, combining sensor data from the TUG, Balance and FTSS tests to a single classifier model per gender yielded mean cross-validated classification accuracy of 87.58% (95% CI: 84.47-91.03%) for the male model and 78.11% (95% CI: 75.38-81.10%) for the female model. These results compared well or exceeded those for classifier models for each test taken individually. Similarly, when classifying frailty status, combining sensor data from the TUG, balance and FTSS tests to a single classifier model per gender, yielded mean cross-validated classification accuracy of 93.94% (95% CI: 91.16-96.51%) for the male model and 84.14% (95% CI: 82.11-86.33%) for the female model (mean 89.04%) which compared well or exceeded results for physical tests taken individually.Results suggest that the combination of these three tests, quantified using body-worn inertial sensors, could lead to improved methods for assessing frailty and falls risk.

Frailty status can be accurately assessed using inertial sensors and the TUG test

BACKGROUND

frailty is an important geriatric syndrome linked to increased mortality, morbidity and falls risk.

METHODS

a total of 399 community-dwelling older adults were assessed using Fried's frailty phenotype and the timed up and go (TUG) test. Tests were quantified using shank-mounted inertial sensors. We report a regression-based method for assessment of frailty using inertial sensor data obtained during TUG. For comparison, frailty was also assessed using the same method based on grip strength and manual TUG time.

RESULTS

using inertial sensor data, participants were classified as frail or non-frail with mean accuracy of 75.20% (stratified by gender). Using TUG time alone, frailty status was classified correctly with mean classification accuracy of 71.82%. Similarly, using grip strength alone, the frailty status was classified correctly with mean classification accuracy of 77.65%. Stratifying sensor data by gender yielded significantly (p<0.05) increased accuracy in classifying frailty when compared with equivalent manual TUG time-based models.

CONCLUSION

results suggest that a simple protocol involving assessment using a well-known mobility test (Timed Up and Go (TUG)) and inertial sensors can be a fast and effective means of automatic, non-expert assessment of frailty.

Early identification of declining balance in higher functioning older adults, an inertial sensor based method

Higher functioning older adults rarely have their balance assessed clinically and as such early decline in balance is not captured. Early identification of declining balance would facilitate earlier intervention and improved management of the ageing process. This study sought to determine if (a) a once off inertial sensor measurement and (b) changes in inertial sensor measurements one year apart can identify declining balance for higher functioning older adults. One hundred and nineteen community dwelling older adults (58 males; 72.5±5.8 years) completed a timed up and go (TUG) instrumented with inertial sensors and the Berg balance scale (BBS) at two time points, one year apart. Temporal and spatio-temporal gait parameters as well as angular velocity and turn parameters were derived from the inertial sensor data. A change in balance from baseline to follow-up was determined by sub-components of the BBS. Changes in inertial sensor parameters from baseline to follow-up demonstrated strong association with balance decline in higher functioning older adults (e.g. mean medial-lateral angular velocity odds ratio=0.2; 95% CI: 0.1-0.5). The area under the Receiver operating characteristic curve (AUC) ranged from 0.8 to 0.9, a marked improvement over change in TUG time alone (AUC 0.6-0.7). Baseline inertial sensor parameters had a similar association with declining balance as age and TUG time. For higher functioning older adults, the change in inertial sensor parameters over time may reflect declining balance. These measures may be useful clinically, to monitor the balance status of older adults and facilitate earlier identification of balance deficits.

Quantitative assessment of multiple sclerosis using inertial sensors and the TUG test

Multiple sclerosis (MS) is a progressive neurological disorder affecting between 2 and 2.5 million people globally. Tests of mobility form part of clinical assessments of MS. Quantitative assessment of mobility using inertial sensors has the potential to provide objective, longitudinal monitoring of disease progression in patients with MS. The mobility of 21 patients (aged 25-59 years, 8 M, 13 F), diagnosed with relapsing-remitting MS was assessed using the Timed up and Go (TUG) test, while patients wore shank-mounted inertial sensors. This exploratory, cross-sectional study aimed to examine the reliability of quantitative measures derived from inertial sensors during the TUG test, in patients with MS. Furthermore, we aimed to determine if disease status (as measured by the Multiple Sclerosis Impact Scale (MSIS-29) and the Expanded Disability Status Score (EDSS)) can be predicted by assessment using a TUG test and inertial sensors. Reliability analysis showed that 32 of 52 inertial sensors parameters obtained during the TUG showed excellent intrasession reliability, while 11 of 52 showed moderate reliability. Using the inertial sensors parameters, regression models of the EDSS and MSIS-29 scales were derived using the elastic net procedure. Using cross validation, an elastic net regularized regression model of MSIS yielded a mean square error (MSE) of 334.6 with 25 degrees of freedom (DoF). Similarly, an elastic net regularized regression model of EDSS yielded a cross-validated MSE of 1.5 with 6 DoF. Results suggest that inertial sensor parameters derived from MS patients while completing the TUG test are reliable and may have utility in assessing disease state as measured using EDSS and MSIS.

A comparison of cross-sectional and prospective algorithms for falls risk assessment

Falls are the most common cause of injury and hospitalization and one of the principal causes of death and disability in older adults worldwide. Accurate identification of patients at risk of falls could lead to timely medical intervention, reducing the incidence of falls related injuries along with associated costs. The current best practice for studies of falls and falls risk recommends the use of prospective follow-up data. However, the majority of studies reporting sensor based methods for assessment of falls risk employ cross-sectional falls data (falls history). The purpose of this study was to compare the performance of sensor based falls risk assessment algorithms derived from cross-sectional (N=909) and prospective (N=259) datasets in terms of false positive rate. The utility of any classification algorithm is clearly limited by a high false positive rate. An estimate of the false positive rate for both cross-sectional and prospective algorithms was determined using an inertial sensor data set of 611 TUG tests from 55 healthy control subjects, with no history of falls. We aimed to determine which falls risk assessment algorithm is more effective at classifying falls risk in healthy control subjects. The cross-sectional algorithm correctly classified 94.11% of tests, while the prospective algorithm, correctly classified 79.38% of tests. Results suggest that sensor based falls risk assessment algorithms generated using cross-sectional falls data, may be more effective than those generated using prospective data in classifying healthy controls and reducing associated false positives.

Falls classification using tri-axial accelerometers during the five-times-sit-to-stand test

The five-times-sit-to-stand test (FTSS) is an established assessment of lower limb strength, balance dysfunction and falls risk. Clinically, the time taken to complete the task is recorded with longer times indicating increased falls risk. Quantifying the movement using tri-axial accelerometers may provide a more objective and potentially more accurate falls risk estimate. 39 older adults, 19 with a history of falls, performed four repetitions of the FTSS in their homes. A tri-axial accelerometer was attached to the lateral thigh and used to identify each sit-stand-sit phase and sit-stand and stand-sit transitions. A second tri-axial accelerometer, attached to the sternum, captured torso acceleration. The mean and variation of the root-mean-squared amplitude, jerk and spectral edge frequency of the acceleration during each section of the assessment were examined. The test-retest reliability of each feature was examined using intra-class correlation analysis, ICC(2,k). A model was developed to classify participants according to falls status. Only features with ICC>0.7 were considered during feature selection. Sequential forward feature selection within leave-one-out cross-validation resulted in a model including four reliable accelerometer-derived features, providing 74.4% classification accuracy, 80.0% specificity and 68.7% sensitivity. An alternative model using FTSS time alone resulted in significantly reduced classification performance. Results suggest that the described methodology could provide a robust and accurate falls risk assessment.

Displacement of centre of mass during quiet standing assessed using accelerometry in older fallers and non-fallers

Postural sway during quiet standing is associated with falls risk in older adults. The aim of this study was to investigate the utility of a range of accelerometer-derived parameters of centre of mass (COM) displacement in identifying older adults at risk of falling. A series of instrumented standing balance trials were performed to investigate postural control in a group of older adults, categorised as fallers or non-fallers. During each trial, participants were asked to stand as still as possible under two conditions: comfortable stance (six repetitions) and semi-tandem stance (three repetitions). A tri-axial accelerometer was secured to the lower back during the trials. Accelerometer data were twice integrated to estimate COM displacement during the trials, with numerical techniques used to reduce integration error. Anterior-posterior (AP) and medial-lateral (ML) sway range, sway length and sway velocity were examined, along with root mean squared (RMS) acceleration. All derived parameters significantly discriminated fallers from non-fallers during both comfortable and semi-tandem stance. Results indicate that these accelerometer-based estimates of COM displacement may improve the discriminative power of quiet standing falls risk assessments, with potential for use in unsupervised balance assessment.

Effects of a low-volume, vigorous intensity step exercise program on functional mobility in middle-aged adults

Aging-related decline in functional mobility is associated with loss of independence. This decline may be mitigated through programs of physical activity. Despite reports of aging-related mobility impairment in middle-aged adults, this age group has been largely overlooked in terms of exercise programs that target functional mobility and the preservation of independence in older age. A method to quantitatively assess changes in functional mobility could direct rehabilitation in a proactive rather than reactive manner. Thirty-three healthy but sedentary middle-aged adults participated in a four week low-volume, vigorous intensity stepping exercise program. Two baseline testing sessions and one post-training testing session were conducted. Functional mobility was assessed using the timed up and go (TUG) test, with its constituent sit-to-walk and walk-to-sit phases examined using a novel inertial sensor-based method. Additionally, semi-tandem balance and knee extensor muscle isometric torque were assessed. Trunk acceleration during walk-to-sit reduced significantly post-training, suggesting altered movement control due to the exercise program. No significant training-induced changes in sit-to-walk acceleration, TUG time, balance or torque were observed. The novel method of functional mobility assessment presented provides a reliable means to quantify subtle changes in mobility during postural transitions. Over time, this exercise program may improve functional mobility.

Quantitative falls risk estimation through multi-sensor assessment of standing balance

Falls are the most common cause of injury and hospitalization and one of the principal causes of death and disability in older adults worldwide. Measures of postural stability have been associated with the incidence of falls in older adults. The aim of this study was to develop a model that accurately classifies fallers and non-fallers using novel multi-sensor quantitative balance metrics that can be easily deployed into a home or clinic setting. We compared the classification accuracy of our model with an established method for falls risk assessment, the Berg balance scale. Data were acquired using two sensor modalities--a pressure sensitive platform sensor and a body-worn inertial sensor, mounted on the lower back--from 120 community dwelling older adults (65 with a history of falls, 55 without, mean age 73.7 ± 5.8 years, 63 female) while performing a number of standing balance tasks in a geriatric research clinic. Results obtained using a support vector machine yielded a mean classification accuracy of 71.52% (95% CI: 68.82-74.28) in classifying falls history, obtained using one model classifying all data points. Considering male and female participant data separately yielded classification accuracies of 72.80% (95% CI: 68.85-77.17) and 73.33% (95% CI: 69.88-76.81) respectively, leading to a mean classification accuracy of 73.07% in identifying participants with a history of falls. Results compare favourably to those obtained using the Berg balance scale (mean classification accuracy: 59.42% (95% CI: 56.96-61.88)). Results from the present study could lead to a robust method for assessing falls risk in both supervised and unsupervised environments.

Body-worn sensor based surrogates of minimum ground clearance in elderly fallers and controls

Falls in the elderly are a major problem worldwide with enormous associated economic and societal costs. Minimum ground clearance (MGC) is an important gait variable when considering trip-related falls risk. This study aimed to investigate the clinical relevance of inertial sensor derived parameters, previously shown to be related to MGC. Previous research by the authors reported a surrogate method for assessing minimum ground clearance (MGC) using shank-mounted inertial sensors in young controls. The present study tests this method on a cohort of 114 community dwelling elderly adults, with and without a history of falls, completing a 30 m continuous walk. Parameters based on the shank angular velocity signals that were shown to be associated with MGC showed significant differences (p<0.05) between fallers and non-fallers yet did not correlate strongly (r<0.7) with two standard measures of falls risk (TUG & BBS). Weak correlations were observed between the angular velocity derived parameters and gait velocity. We conclude that these parameters are clinically meaningful and therefore may constitute a new measure of falls risk.

An instrumented sit-to-stand test used to examine differences between older fallers and non-fallers

An instrumented version of the five-times-sit-to-stand test was performed in the homes of a group of older adults, categorised as fallers or non-fallers. Tri-axial accelerometers were secured to the sternum and anterior thigh of each participant during the assessment. Accelerometer data were then used to examine the timing of the movement, as well as the root mean squared amplitude, jerk and spectral edge frequency of the mediolateral (ML) acceleration during the total assessment, each sit-stand-sit component and each postural transition (sit-stand and stand-sit). Differences between fallers and non-fallers were examined for each parameter. Six parameters significantly discriminated between fallers and non-fallers: sit-stand time, ML acceleration for the total assessment, and the ML spectral edge frequency for the complete assessment, individual sit-stand-sit components, as well as sit-stand and stand-sit transitions. These results suggest that each of these derived parameters would provide improved discrimination of fallers from non-fallers, for the cohort examined, than the standard clinical measure - the total time to complete the assessment. These results indicate that accelerometry may enhance the utility of the five-times-sit-to-stand test when assessing falls risk.

Development and validation of a clinic based balance assessment technology

Falls in the elderly are a major problem worldwide, with enormous associated societal costs. Deficits in balance and postural control have long been associated with falls risk in elderly adults. The gold standard for quantitative assessment of human balance in a clinical setting is the force plate which is highly expensive, non-portable and requires specialized personnel to operate. The present study aims to evaluate the validity and reliability of a portable quantitative balance measurement technology compared to the forceplate. Two participants (1 male, 1 female) performed sixteen balance trials each (eight eyes open and eight eyes closed). Simultaneous data were recorded from a portable pressure sensor platform and a laboratory grade force platform. Standard centre of pressure (COP) metrics from both modalities were compared and high levels of agreement in terms of intraclass correlation coefficient (ICC), mean absolute error (MAE) and mean percentage error (MPE) were found.

Reliability of quantitative TUG measures of mobility for use in falls risk assessment

Recent advances in body-worn sensor technology have increased the scope for harnessing quantitative information from the timed-up-and-go test (TUG), well beyond simply the time taken to perform the test. Previous research has shown that the quantitative TUG method can differentiate fallers from non-fallers with greater success than the manually timed TUG or the Berg Balance Test. In order to advance this paradigm of falls risk estimation it is necessary to investigate the robustness of the quantitative TUG variables. This study investigated the inter-session and intra-session reliability of 44 quantitative TUG variables measured from the shanks and lower back of 33 study participants aged between 55-65 yrs. For intra-session reliability, 25 variables demonstrated excellent reliability (ICC>0.75), and 12 demonstrated "fair to good reliability" with ICCs between 0.4 and 0.75. Analysis of test-retest reliability resulted in ICC > 0.75 for 18 out of 44 variables, with 20 variables showing fair to good reliability. Turn time parameters demonstrated poor reliability. We conclude that this is a reliable instrument that may be used as part of a long-term falls risk assessment, with further work required to improve certain turn parameters.

Diurnal variations in the outcomes of instrumented gait and quiet standing balance assessments and their association with falls history

One in three adults aged over 65 falls every year, resulting in enormous costs to society. Incidents of falling vary with time of day, peaking in the early morning. The aim of this study was to determine if the ability of instrumented gait and balance assessments to discriminate between participants based on their falls history varies diurnally. Body-worn sensors were used during a 3 m gait assessment and a series of quiet standing balance tests. Each assessment was performed four times during a single day under supervised conditions in the participant's homes. 40 adults aged over 60 years (19 fallers) participated in this study. A range of parameters were derived for each assessment, and the ability of each parameter to discriminate between fallers and non-fallers at each recording time was examined. The effect of falls history on single support time varied significantly with recording time, with a significantly reduced single support time observed at the first and last recording session of the day. Differences were observed between fallers and non-fallers for a range of other gait parameters; however, these effects did not vary with assessment time. The quiet standing assessments examined in this study revealed significant variations with falls history; however, the sensitivity of the examined quiet standing assessments to falls risk does not appear to be time dependent. These results indicate that, with the exception of single support time, the association of gait and quiet standing balance parameters with falls risk does not vary diurnally.

Taking balance measurement out of the laboratory and into the home: discriminatory capability of novel centre of pressure measurement in fallers and non-fallers

We investigated three methods for estimating centre of pressure excursions, as measured using a portable pressure sensor matrix, in order to deploy similar technology into the homes of older adults for longitudinal monitoring of postural control and falls risk. We explored the utility of these three methods as markers of falls risk in a cohort of 120 community dwelling older adults with and without a history of falls (65 fallers, 55 non-fallers). A number of standard quantitative balance parameters were derived using each centre of pressure estimation method. Rank sum tests were used to test for significant differences between fallers and non-fallers while intra-class correlation coefficients were also calculated to determine the reliability of each method. A method based on estimating the changes in the magnitude of pressure exerted on the pressure sensor matrix was found to be the most reliable and discriminative. Our future work will implement this method for home-based balance measurement.

Technology Innovation Enabling Falls Risk Assessment in a Community Setting

Approximately one in three people over the age of 65 will fall each year, resulting in significant financial, physical, and emotional cost on the individual, their family, and society. Currently, falls are managed using on-body sensors and alarm pendants to notify others when a falls event occurs. However these technologies do not prevent a fall from occurring. There is now a growing focus on falls risk assessment and preventative interventions. Falls risk is currently assessed in a clinical setting by expert physiotherapists, geriatricians, or occupational therapists following the occurrence of an injurious fall. As the population ages, this reactive model of care will become increasingly unsatisfactory, and a proactive community-based prevention strategy will be required. Recent advances in technology can support this new model of care by enabling community-based practitioners to perform tests that previously required expensive technology or expert interpretation. Gait and balance impairment is one of the most common risk factors for falls. This paper reviews the current technical and non-technical gait and balance assessments, discusses how low-cost technology can be applied to objectively administer and interpret these tests in the community, and reports on recent research where body-worn sensors have been utilized. It also discusses the barriers to adoption in the community and proposes ethnographic research as a method to investigate solutions to these barriers.

A single gyroscope method for spatial gait analysis

Inertial sensors have become increasingly popular in gait analysis, due to their highly portable, low cost, and potentially wireless nature. However, accurate spatial gait analysis using few sensors remains a challenge. A gyroscope-based algorithm for spatial gait analysis is presented. This novel algorithm (SGA) uses data from a single gyroscope attached to each shank. The performance of the SGA was compared to that of an electronic walkway, GAITRite®. The two systems compared favorably, with a mean error in stride length of 0.09 ± 0.07 m, and a mean error in stride velocity of 0.11 ± 0.10 m/s. The error between the SGA and GAITRite was also similar to that reported by previous inertial sensor based algorithms. The relationship between stride length and stride velocity, as well as that of subject height and stride length was also examined. This new method provides an inexpensive, portable system for spatial or spatio-temporal gait analysis, which has potential for use in any location.

SHIMMER: an extensible platform for physiological signal capture

Wireless sensor networks have become increasingly common in everyday applications due to decreasing technology costs and improved product performance, robustness and extensibility. Wearable physiological monitoring systems have been utilized in a variety of studies, particularly those investigating ECG or EMG during human movement or sleep monitoring. These systems require extensive validation to ensure accurate and repeatable functionality. Here we validate the physiological signals (EMG, ECG and GSR) of the SHIMMER (Sensing Health with Intelligence, Modularity, Mobility and Experimental Reusability) against known commercial systems. Signals recorded by the SHIMMER EMG, ECG and GSR daughter-boards were found to compare well to those obtained by commercial systems.

Clinical Gait assessment of older adults using open platform tools

Gait impairment is associated with increased falls risk. The gait of 321 community dwelling elderly adults was assessed using the TRIL Gait Analysis Platform (GAP), which was specially designed for ease of use in a research clinic setting by non-experts. The GAP featured body-worn kinematic sensors, a pressure sensitive electronic walkway, and two orthogonally mounted web cameras, and was developed using open platform tools. This flexible platform was applied to objectively measure gait parameters in different gait assessments. The results from the 6 meter walk assessment are presented here. In this assessment, participants were categorized by clinical falls history as 'fallers' or 'non-fallers'. Temporal and spatial gait parameters were examined. Significant differences in spatial parameters were observed when fallers and non-fallers were compared. Temporal parameters were found to differ, though not significantly.

Adaptive estimation of temporal gait parameters using body-worn gyroscopes

Body-worn kinematic sensors have been widely proposed for use in portable, low cost, ambulatory monitoring of gait. Such sensor based systems could avoid the need for high-cost laboratory-based methods for measurement of gait. We aimed to evaluate an adaptive gyroscope-based algorithm for automated temporal gait analysis using body-worn wireless gyroscopes. Temporal gait parameters were calculated from initial contact (IC) and terminal contact (TC) points derived from gyroscopes, contained in wireless sensors on the left and right shanks, using a newly developed adaptive algorithm. Gyroscope data from nine healthy adult subjects performing four walks at three different speeds were then compared against data acquired simultaneously using two force-plates. Results show that the mean true error between the adaptive gyroscope algorithm and force-plate was -5.5 ± 7.3 ms and 40.6 ± 19.2 ms for IC and TC points respectively; the latter representing a consistent, systematic error of this magnitude that may be intrinsic to shank-mounted gyroscopes. These results suggest that the algorithm reported here could form the basis of a robust, portable, low-cost system for ambulatory monitoring of gait.

An adaptive gyroscope-based algorithm for temporal gait analysis

Body-worn kinematic sensors have been widely proposed as the optimal solution for portable, low cost, ambulatory monitoring of gait. This study aims to evaluate an adaptive gyroscope-based algorithm for automated temporal gait analysis using body-worn wireless gyroscopes. Gyroscope data from nine healthy adult subjects performing four walks at four different speeds were then compared against data acquired simultaneously using two force plates and an optical motion capture system. Data from a poliomyelitis patient, exhibiting pathological gait walking with and without the aid of a crutch, were also compared to the force plate. Results show that the mean true error between the adaptive gyroscope algorithm and force plate was -4.5 ± 14.4 ms and 43.4 ± 6.0 ms for IC and TC points, respectively, in healthy subjects. Similarly, the mean true error when data from the polio patient were compared against the force plate was -75.61 ± 27.53 ms and 99.20 ± 46.00 ms for IC and TC points, respectively. A comparison of the present algorithm against temporal gait parameters derived from an optical motion analysis system showed good agreement for nine healthy subjects at four speeds. These results show that the algorithm reported here could constitute the basis of a robust, portable, low-cost system for ambulatory monitoring of gait.

Quantitative falls risk assessment using the timed up and go test

Falls are a major problem in older adults worldwide with an estimated 30% of elderly adults over 65 years of age falling each year. The direct and indirect societal costs associated with falls are enormous. A system that could provide an accurate automated assessment of falls risk prior to falling would allow timely intervention and ease the burden on overstretched healthcare systems worldwide. An objective method for assessing falls risk using body-worn kinematic sensors is reported. The gait and balance of 349 community-dwelling elderly adults was assessed using body-worn sensors while each patient performed the "timed up and go" (TUG) test. Patients were also evaluated using the Berg balance scale (BBS). Of the 44 reported parameters derived from body-worn kinematic sensors, 29 provided significant discrimination between patients with a history of falls and those without. Cross-validated estimates of retrospective falls prediction performance using logistic regression models yielded a mean sensitivity of 77.3% and a mean specificity of 75.9%. This compares favorably to the cross-validated performance of logistic regression models based on the time taken to complete the TUG test (manually timed TUG) and the Berg balance score. These models yielded mean sensitivities of 58.0% and 57.8%, respectively, and mean specificities of 64.8% and 64.2%, respectively. Results suggest that this method offers an improvement over two standard falls risk assessments (TUG and BBS) and may have potential for use in supervised assessment of falls risk as part of a longitudinal monitoring protocol.

Objective real-time assessment of walking and turning in elderly adults

Recent research suggests that falls are the most common cause of injury and disability in older persons. Invasive systems or body worn sensors can be employed in controlled clinical and laboratory settings to determine clinical measures of gait and stability. This study by contrast aims to explore how video technology, can be employed to unobtrusively determine the same measures. Data from 63 elderly subjects, recruited through a research clinic was analyzed. The derived parameters include: the walk time, the number of steps of the TUG test and stability out of the turn. The results show that video analysis can be used to automate current clinical measures of gait and stability as well as to inform future automated interventions.

Characterisation of heart rate changes and their correlation with EEG during neonatal seizures

The effect of seizures on instantaneous HR (iHR) in 12 neonates is investigated here. HR can be readily extracted from the ECG and can be employed as an additional signal in seizure detection algorithms. The change in instantaneous HR and its correlation with the change in RMS EEG amplitude were examined. Two methods were employed to classify significant iHR changes. Significant correlation (p 0.05) during seizure was observed in 100% of patients (83.33% of seizures). Overall, significant iHR changes (classified by either method) were found in 83% of patients (50% of seizures). It was found that a markedly higher iHR was observed in patients whose seizures were not classified as having significant iHR changes.

SHIMMER: A new tool for temporal gait analysis

Development of a flexible wireless sensor platform for measurement of biomechanical and physiological variables related to functional movement would be a vital step towards effective ambulatory monitoring and early detection of risk factors in the ageing population. The small form factor, wirelessly enabled SHIMMER platform has been developed towards this end. This study is focused assessing the utility of the SHIMMER for use in ambulatory human gait analysis. Temporal gait parameters derived from a tri-axial gyroscope contained in the SHIMMER are compared against those acquired simultaneously using the CODA motion analysis system. Results from a healthy adult male subject show excellent agreement (ICC(2, k) > 0.85) in stride, swing and stance time for 10 walking trials and 4 run trials. The mean differences using the Bland and Altman method for stance, stride and swing times were 0.0087, 0.0044 and -0.0061 seconds respectively. These results suggest that the SHIMMER is a versatile cost effective tool for use in temporal gait analysis.

A comparison of quantitative EEG features for neonatal seizure detection

OBJECTIVE

This study was undertaken to identify the best performing quantitative EEG features for neonatal seizures detection from a test set of 21.

METHODS

Each feature was evaluated on 1-min, artefact-free segments of seizure and non-seizure neonatal EEG recordings. The potential utility of each feature for neonatal seizure detection was determined using receiver operating characteristic analysis and repeated measures t-tests. A performance estimate of the feature set was obtained using a cross-fold validation and combining all features together into a linear discriminant classifier model.

RESULTS

Significant differences between seizure and non-seizure segments were found in 19 features for 17 patients. The best performing features for this application were the RMS amplitude, the line length and the number of local maxima and minima. An estimate of the patient independent classifier performance yielded a sensitivity of 81.08% and specificity of 82.23%.

CONCLUSIONS

The individual performances of 21 quantitative EEG features in detecting electrographic seizure in the neonate were compared and numerically quantified. Combining all features together into a classifier model led to superior performance than that provided by any individual feature taken alone.

SIGNIFICANCE

The results documented in this study may provide a reference for the optimum quantitative EEG features to use in developing and enhancing neonatal seizure detection algorithms.

Seizure detection in neonates: Improved classification through supervised adaptation

The goal of neonatal seizure detection is the development of a patient independent system to alert staff in the neonatal intensive care unit of ongoing seizures. This study demonstrates the potential in adapting a patient independent classifier using patient specific data. Supervised adaptation is investigated using the basic gradient descent algorithm and least mean squares procedures. An increase in mean ROC area of 3% is obtained for the best performing learning algorithm, yielding an increase in mean accuracy of 7.7% compared to the patient independent algorithm.

Automated single channel seizure detection in the neonate

Neonatal seizures are the most common neurological emergency in the neonatal period and are associated with poor long-term outcome. EEG is considered the gold standard for identification of all neonatal seizures, reducing the number of EEG electrodes required would reduce patient handling and allow faster acquisition of data. A method for automated neonatal seizure detection based on two carefully chosen cerebral scalp electrodes but trained using multi-channel EEG is presented. The algorithm was developed and tested using a multi-channel EEG dataset containing 411 seizures from 251.9 hours of EEG recorded from 17 full-term neonates. Automated seizure detection using a variety of bipolar channel derivations was investigated. Channel C3-C4 yielded correct detection of 90.77% of seizures with a false detection rate of 9.43%. This compares favourably with a multi-channel seizure detection method which detected 81.03% of seizures with a false detection rate of 3.82%.

How often are physicians and chiropractors provided with patient information when accepting referrals?

The objective of our study is to document how often MD/DOs and doctors of chiropractic (DCs) receive patient information from referring MD/DOs and DCs and highlight to what extent there is a lack of formal intraprofessional and interprofessional referral relationships between MD/DOs and DCs. A total of 517 MD/DOs and 452 DCs participated in this study. The study results suggest that patient information is not regularly provided by either MD/DOs or DCs, even when making formal referrals to a provider of the same type. This was more pronounced when MDs made formal referrals to DCs.

Reinventing CME: the role of the care pilot in the medical group practice

This article recommends that the content of traditional continuing medical education be changed significantly to include the concepts and skills necessary to enable practice teams to feedback information into the practice, which would result in the creation of a learning organization with the ability to plan for and anticipate future activities. The primary role in this new organization would be called a care pilot who would have as a primary responsibility, the successful navigation and improvement of the 6 aims as spelled out in the Institute of Medicine report Crossing the Quality Chasm.

Facilitators and Barriers to Improving Interprofessional Referral Relationships Between Primary Care Physicians and Chiropractors

OBJECTIVE

Findings from recent studies suggest that there are poor interprofessional referral relationships between primary care physicians (MDs) and chiropractors (DCs) and this can lead to fragmentation of care. The objective of this study is to identify potential facilitators and barriers to developing positive interprofessional referrals relationships between MDs and DCs.

METHODS

We conducted 2 rounds of focus group interviews on a convenience sample of MDs and DCs. The focus groups were audiotaped, and transcripts were prepared for each focus group interaction. These data were analyzed through content analysis by 2 independent evaluators to determine the key themes and concepts provided by the focus groups.

RESULTS

Both MDs and DCs suggested that good communication, openness to discussion by providers, and patient interest are key factors for developing positive interprofessional referral relationships and implementing interprofessional practice-based research networks. Barriers to interprofessional relationships include lack of good communication between the 2 provider types, bias toward alternative medicine, lack of knowledge or understanding of chiropractic care, geographic constraints, and economic considerations.

CONCLUSIONS

This study identified several facilitators and barriers for developing positive referral relationships between primary care physicians and chiropractors. Future studies must focus on demonstrating the role of these factors on developing positive interprofessional relationships.