Validity of using tri-axial accelerometers to measure human movement - Part II: Step counts at a wide range of gait velocities

Mechanical Method for Rapid Determination of Step Count Sensor Settings

With the increased push for personalized medicine, researchers and clinicians have begun exploring the use of wearable sensors to track patient activity. These sensors typically prioritize device life over robust onboard analysis, which results in lower accuracies in step count, particularly at lower cadences. To optimize the accuracy of activity-monitoring devices, particularly at slower walking speeds, proven methods must be established to identify suitable settings in a controlled and repeatable manner prior to human validation trials. Currently, there are no methods for optimizing these low-power wearable sensor settings prior to human validation, which requires manual counting for in-laboratory participants and is limited by time and the cadences that can be tested. This article proposes a novel method for determining sensor step counting accuracy prior to human validation trials by using a mechanical camshaft actuator that produces continuous steps. Sensor error was identified across a representative subspace of possible sensor setting combinations at cadences ranging from 30 steps/min to 110 steps/min. These true errors were then used to train a multivariate polynomial regression to model errors across all possible setting combinations and cadences. The resulting model predicted errors with an R2 of 0.8 and root-mean-square error (RMSE) of 0.044 across all setting combinations. An optimization algorithm was then used to determine the combinations of settings that produced the lowest RMSE and median error for three ranges of cadence that represent disabled low-mobility ambulators, disabled high-mobility ambulators, and healthy ambulators (30-60, 20-90, and 30-110 steps/min, respectively). The model identified six setting combinations for each range of interest that achieved a ±10% error in cadence prior to human validation. The anticipated range of errors from the optimized settings at lower walking speeds are lower than the reported errors of wearable sensors (±30%), suggesting that pre-human-validation optimization of sensors may decrease errors at lower cadences. This method provides a novel and efficient approach to optimizing the accuracy of wearable activity monitors prior to human validation trials.

Comparison of the effects of rhythmic vibrotactile stimulations and rhythmic auditory stimulations on Parkinson's disease patients' gait variability: a pilot study

INTRODUCTION

Parkinson's disease patients' gait is characterized by shorter step length, reduced gait velocity and deterioration of temporal organization of stride duration variability (modified Long Range Autocorrelations). The objective of this study was to compare effects of rhythmic auditory stimulations (RAS) and Rhythmic Vibrotactile Stimulations (RVS) on Parkinson's disease patients' gait.

METHODS

Ten Parkinson's disease patients performed three walking conditions lasting 5-7 min each: control condition (CC), RAS condition and RVS condition. Inertial measurement units were used to assess spatiotemporal gait parameters. Stride duration variability was assessed in terms of magnitude using coefficient of variation and in terms of temporal organization (i.e., Long Range Autocorrelations computation) using the evenly spaced averaged Detrended Fluctuation Analysis (α-DFA exponent).

RESULTS

Gait velocity was significantly higher during RAS condition than during CC (Cohen's d = 0.52) and similar to RVS condition (Cohen's d = 0.17). Cadence was significantly higher during RAS (Cohen's d = 0.77) and RVS (Cohens' d = 0.56) conditions than during CC. Concerning variability, no difference was found either for mean coefficient of variation or mean α-DFA between conditions. However, a great variability of individual results between the RAS and the RVS conditions is to be noted concerning α-DFA.

CONCLUSIONS

RAS and RVS improved similarly PD patients' spatiotemporal gait parameters, without modifying stride duration variability in terms of magnitude and temporal organization at group level. Future studies should evaluate the relevant parameters for administering the right cueing type for the right patient.

TRIAL REGISTRATION

ClinicalTrial.gov registration number NCT05790759, date of registration: 16/03/2023, retrospectively registered.

Person-Specific Template Matching Using a Dynamic Time Warping Step-Count Algorithm for Multiple Walking Activities

This study aimed to develop and evaluate a new step-count algorithm, StepMatchDTWBA, for the accurate measurement of physical activity using wearable devices in both healthy and pathological populations. We conducted a study with 30 healthy volunteers wearing a wrist-worn MOX accelerometer (Maastricht Instruments, NL). The StepMatchDTWBA algorithm used dynamic time warping (DTW) barycentre averaging to create personalised templates for representative steps, accounting for individual walking variations. DTW was then used to measure the similarity between the template and accelerometer epoch. The StepMatchDTWBA algorithm had an average root-mean-square error of 2 steps for healthy gaits and 12 steps for simulated pathological gaits over a distance of about 10 m (GAITRite walkway) and one flight of stairs. It outperformed benchmark algorithms for the simulated pathological population, showcasing the potential for improved accuracy in personalised step counting for pathological populations. The StepMatchDTWBA algorithm represents a significant advancement in accurate step counting for both healthy and pathological populations. This development holds promise for creating more precise and personalised activity monitoring systems, benefiting various health and wellness applications.

Outcome of bracing vs. surgical treatment in adolescents with idiopathic scoliosis based on device measured daily physical activity: a prospective pilot study

Adolescent idiopathic scoliosis (AIS) can be treated with bracing or surgery, which may affect patient's physical activity (PA). However, there are limited objective assessments of PA in patients with AIS. This study aims to compare the outcome of spinal bracing vs. surgery in patients with AIS based on a device that measured daily PA. In total 24 patients with AIS participated, including 12 patients treated with bracing and 12 with spinal surgery. Daily PA was measured throughout 4 consecutive days using four tri-axial accelerometers and patient-reported functional status was reported using the SRS-22 questionnaire. The participants were assessed both before the treatment and after treatment at a 12-month follow-up. Patients with AIS had no significant change in their PA levels at the 12-month follow-up after surgical correction. On the contrary, patients with AIS following a year-long bracing treatment had significantly reduced time spent active ( P  = 0.04) with an average reduction in walking steps by 2137 steps/day ( P  = 0.005). There was no significant difference in function, pain, self-image and mental health domains following both treatments, as reported by the SRS-22. There was a significant improvement in satisfaction for both treatment groups ( P  ≤ 0.02). Significantly reduced PA and increased sedentary time are reported in patients with AIS following bracing treatment. An objective PA assessment is recommended to track the effect of scoliosis treatment on PA. Patients with AIS should be actively encouraged to achieve and maintain their recommended daily PA levels irrespective of the type of treatment. Level of evidence: Level II.

Relationship of Daily Step Counts to All-Cause Mortality and Cardiovascular Events

BACKGROUND

The minimal and optimal daily step counts for health improvements remain unclear.

OBJECTIVES

A meta-analysis was performed to quantify dose-response associations of objectively measured step count metrics in the general population.

METHODS

Electronic databases were searched from inception to October 2022. Primary outcomes included all-cause mortality and incident cardiovascular disease (CVD). Study results were analyzed using generalized least squares and random-effects models.

RESULTS

In total, 111,309 individuals from 12 studies were included. Significant risk reductions were observed at 2,517 steps/d for all-cause mortality (adjusted HR [aHR]: 0.92; 95% CI: 0.84-0.999) and 2,735 steps/d for incident CVD (aHR: 0.89; 95% CI: 0.79-0.999) compared with 2,000 steps/d (reference). Additional steps resulted in nonlinear risk reductions of all-cause mortality and incident CVD with an optimal dose at 8,763 (aHR: 0.40; 95% CI: 0.38-0.43) and 7,126 steps/d (aHR: 0.49; 95% CI: 0.45-0.55), respectively. Increments from a low to an intermediate or a high cadence were independently associated with risk reductions of all-cause mortality. Sex did not influence the dose-response associations, but after stratification for assessment device and wear location, pronounced risk reductions were observed for hip-worn accelerometers compared with pedometers and wrist-worn accelerometers.

CONCLUSIONS

As few as about 2,600 and about 2,800 steps/d yield significant mortality and CVD benefits, with progressive risk reductions up to about 8,800 and about 7,200 steps/d, respectively. Additional mortality benefits were found at a moderate to high vs a low step cadence. These findings can extend contemporary physical activity prescriptions given the easy-to-understand concept of step count. (Dose-Response Relationship Between Daily Step Count and Health Outcomes: A Systematic Review and Meta-Analyses; CRD42021244747).

Waist-What? Can a single sensor positioned at the waist detect parameters of gait at a speed and distance reflective of older adults' activity?

One of the problems facing an ageing population is functional decline associated with reduced levels of physical activity (PA). Traditionally researcher or clinician input is necessary to capture parameters of gait or PA. Enabling older adults to monitor their activity independently could raise their awareness of their activitiy levels, promote self-care and potentially mitigate the risks associated with ageing. The ankle is accepted as the optimum position for sensor placement to capture parameters of gait however, the waist is proposed as a more accessible body-location for older adults. This study aimed to compare step-count measurements obtained from a single inertial sensor positioned at the ankle and at the waist to that of a criterion measure of step-count, and to compare gait parameters obtained from the sensors positioned at the two different body-locations. Step-count from the waist-mounted inertial sensor was compared with that from the ankle-mounted sensor, and with a criterion measure of direct observation in healthy young and healthy older adults during a three-minute treadmill walk test. Parameters of gait obtained from the sensors at both body-locations were also compared. Results indicated there was a strong positive correlation between step-count measured by both the ankle and waist sensors and the criterion measure, and between ankle and waist sensor step-count, mean step time and mean stride time (r = .802-1.0). There was a moderate correlation between the step time variability measures at the waist and ankle (r = .405). This study demonstrates that a single sensor positioned at the waist is an appropriate method for the capture of important measures of gait and physical activity among older adults.

Fourteen days free-living evaluation of an open-source algorithm for counting steps in healthy adults with a large variation in physical activity level

BACKGROUND

The number of steps by an individual, has traditionally been assessed with a pedometer, but increasingly with an accelerometer. The ActiLife software (AL) is the most common way to process accelerometer data to steps, but it is not open source which could aid understanding of measurement errors. The aim of this study was to compare assessment of steps from the open-source algorithm part of the GGIR package and two closed algorithms, AL normal (n) and low frequency extension (lfe) algorithms to Yamax pedometer, as reference. Free-living in healthy adults with a wide range of activity level was studied.

RESULTS

A total 46 participants divided by activity level into a low-medium active group and a high active group, wore both an accelerometer and a pedometer for 14 days. In total 614 complete days were analyzed. A significant correlation between Yamax and all three algorithms was shown but all comparisons were significantly different with paired t-tests except for ALn vs Yamax. The mean bias shows that ALn slightly overestimated steps in the low-medium active group and slightly underestimated steps in high active group. The mean percentage error (MAPE) was 17% and 9% respectively. The ALlfe overestimated steps by approximately 6700/day in both groups and the MAPE was 88% in the low-medium active group and 43% in the high active group. The open-source algorithm underestimated steps with a systematic error related to activity level. The MAPE was 28% in the low-medium active group and 48% in the high active group.

CONCLUSION

The open-source algorithm captures steps fairly well in low-medium active individuals when comparing with Yamax pedometer, but did not show satisfactory results in more active individuals, indicating that it must be modified before implemented in population research. The AL algorithm without the low frequency extension measures similar number of steps as Yamax in free-living and is a useful alternative before a valid open-source algorithm is available.

Development and Validation of a Machine Learning Wrist-worn Step Detection Algorithm with Deployment in the UK Biobank

Background

Step count is an intuitive measure of physical activity frequently quantified in a range of health-related studies; however, accurate quantification of step count can be difficult in the free-living environment, with step counting error routinely above 20% in both consumer and research-grade wrist-worn devices. This study aims to describe the development and validation of step count derived from a wrist-worn accelerometer and to assess its association with cardiovascular and all-cause mortality in a large prospective cohort study.

Methods

We developed and externally validated a hybrid step detection model that involves self-supervised machine learning, trained on a new ground truth annotated, free-living step count dataset (OxWalk, n=39, aged 19-81) and tested against other open-source step counting algorithms. This model was applied to ascertain daily step counts from raw wrist-worn accelerometer data of 75,493 UK Biobank participants without a prior history of cardiovascular disease (CVD) or cancer. Cox regression was used to obtain hazard ratios and 95% confidence intervals for the association of daily step count with fatal CVD and all-cause mortality after adjustment for potential confounders.

Findings

The novel step algorithm demonstrated a mean absolute percent error of 12.5% in free-living validation, detecting 98.7% of true steps and substantially outperforming other recent wrist-worn, open-source algorithms. Our data are indicative of an inverse dose-response association, where, for example, taking 6,596 to 8,474 steps per day was associated with a 39% [24-52%] and 27% [16-36%] lower risk of fatal CVD and all-cause mortality, respectively, compared to those taking fewer steps each day.

Interpretation

An accurate measure of step count was ascertained using a machine learning pipeline that demonstrates state-of-the-art accuracy in internal and external validation. The expected associations with CVD and all-cause mortality indicate excellent face validity. This algorithm can be used widely for other studies that have utilised wrist-worn accelerometers and an open-source pipeline is provided to facilitate implementation.

A novel approach for modelling and classifying sit-to-stand kinematics using inertial sensors

Sit-to-stand transitions are an important part of activities of daily living and play a key role in functional mobility in humans. The sit-to-stand movement is often affected in older adults due to frailty and in patients with motor impairments such as Parkinson's disease leading to falls. Studying kinematics of sit-to-stand transitions can provide insight in assessment, monitoring and developing rehabilitation strategies for the affected populations. We propose a three-segment body model for estimating sit-to-stand kinematics using only two wearable inertial sensors, placed on the shank and back. Reducing the number of sensors to two instead of one per body segment facilitates monitoring and classifying movements over extended periods, making it more comfortable to wear while reducing the power requirements of sensors. We applied this model on 10 younger healthy adults (YH), 12 older healthy adults (OH) and 12 people with Parkinson's disease (PwP). We have achieved this by incorporating unique sit-to-stand classification technique using unsupervised learning in the model based reconstruction of angular kinematics using extended Kalman filter. Our proposed model showed that it was possible to successfully estimate thigh kinematics despite not measuring the thigh motion with inertial sensor. We classified sit-to-stand transitions, sitting and standing states with the accuracies of 98.67%, 94.20% and 91.41% for YH, OH and PwP respectively. We have proposed a novel integrated approach of modelling and classification for estimating the body kinematics during sit-to-stand motion and successfully applied it on YH, OH and PwP groups.

Field based assessment of a tri-axial accelerometers validity to identify steps and reliability to quantify external load

Background: The monitoring of accelerometry derived load has received increased attention in recent years. However, the ability of such measures to quantify training load during sport-related activities is not well established. Thus, the current study aimed to assess the validity and reliability of tri-axial accelerometers to identify step count and quantify external load during several locomotor conditions including walking, jogging, and running.

Method: Thirty physically active college students (height = 176.8 ± 6.1 cm, weight = 82.3 ± 12.8 kg) participated. Acceleration data was collected via two tri-axial accelerometers (Device A and B) sampling at 100 Hz, mounted closely together at the xiphoid process. Each participant completed two trials of straight-line walking, jogging, and running on a 20 m course. Device A was used to assess accelerometer validity to identify step count and the test-retest reliability of the instrument to quantify the external load. Device A and Device B were used to assess inter-device reliability. The reliability of accelerometry-derived metrics Impulse Load (IL) and Magnitude g (MAG) were assessed.

Results: The instrument demonstrated a positive predictive value (PPV) ranging between 96.98%–99.41% and an agreement ranging between 93.08%–96.29% for step detection during all conditions. Good test-retest reliability was found with a coefficient of variation (CV) <5% for IL and MAG during all locomotor conditions. Good inter-device reliability was also found for all locomotor conditions (IL and MAG CV < 5%).

Conclusion: This research indicates that tri-axial accelerometers can be used to identify steps and quantify external load when movement is completed at a range of speeds.

Stop motion: using high resolution spatiotemporal data to estimate and locate stationary and movement behaviour in an office workplace

Prolonged periods of stationary behaviour, a common occurrence in many office workplaces, are linked with a range of physical disorders. Investigating the physical context of this behaviour may be a key to developing effective interventions. This study aimed to estimate and locate the stationary and movement behaviours of office workers (n = 10) by segmenting spatiotemporal data collected over 5 days in an office work-based setting. The segmentation method achieved a balanced accuracy ≥85.5% for observation classification and ≥90% for bout classification when compared to reference data. The results show the workers spent the majority of their time stationary (Mean = 86.4%) and had on average, 28.4 stationary and 25.9 moving bouts per hour. While these findings accord with other studies, the segmented data was also visualised, revealing that the workers were stationary for periods ≥5 min at multiple locations and these locations changed across time. Practitioner Summary: This study applied a data segmentation method to classify stationary and moving behaviours from spatiotemporal data collected in an office workplace. The segmented data revealed not only what behaviours occurred but also their location, duration, and time. Segmenting spatiotemporal data may add valuable physical context to aid workplace research.

Estimation of manual wheelchair-based activities in the free-living environment using a neural network model with inertial body-worn sensors

Shoulder pain is common in manual wheelchair (MWC) users. Overuse is thought to be a major cause, but little is known about exposure to activities of daily living (ADLs). The study goal was to develop a method to estimate three conditions in the field: (1) non-propulsion activity, (2) MWC propulsion, and (3) static time using an inertial measurement unit (IMU). Upper arm IMU data were collected as ten MWC users performed lab-based MWC-related ADLs. A neural network model was developed to classify data as non-propulsion activity, propulsion, or static, and validated for the lab-based data collection by video comparison. Six of the participants' free-living IMU data were collected and the lab-based model was applied to estimate daily non-propulsion activity, propulsion, and static time. The neural network model yielded lab-based validity measures ≥0.87 for differentiating non-propulsion activity, propulsion, and static time. A quasi-validation of one participant's field-based data yielded validity measures ≥0.66 for identifying propulsion. Participants' estimated mean daily non-propulsion activity, propulsion, and static time ranged from 158 to 409, 13 to 25, and 367 to 609 min, respectively. The preliminary results suggest the model may be able to accurately identify MWC users' field-based activities. The inclusion of field-based IMU data in the model could further improve field-based classification.

Skeletal muscle mitochondrial dysfunction and muscle and whole body functional deficits in cancer patients with weight loss

Reductions in skeletal muscle mass and function are often reported in patients with cancer-associated weight loss and are associated with reduced quality of life, impaired treatment tolerance, and increased mortality. Although cellular changes, including altered mitochondrial function, have been reported in animals, such changes have been incompletely characterized in humans with cancer. Whole body and skeletal muscle physical function, skeletal muscle mitochondrial function, and whole body protein turnover were assessed in eight patients with cancer-associated weight loss (10.1 ± 4.2% body weight over 6-12 mo) and 19 age-, sex-, and body mass index (BMI)-matched healthy controls to characterize skeletal muscle changes at the whole body, muscle, and cellular level. Potential pathways involved in cancer-induced alterations in metabolism and mitochondrial function were explored by interrogating skeletal muscle and plasma metabolomes. Despite similar lean mass compared with control participants, patients with cancer exhibited reduced habitual physical activity (57% fewer daily steps), cardiorespiratory fitness [22% lower V̇o_2peak (mL/kg/min)] and leg strength (35% lower isokinetic knee extensor strength), and greater leg neuromuscular fatigue (36% greater decline in knee extensor torque). Concomitant with these functional declines, patients with cancer had lower mitochondrial oxidative capacity [25% lower State 3 O₂ flux (pmol/s/mg tissue)] and ATP production [23% lower State 3 ATP production (pmol/s/mg tissue)] and alterations in phospholipid metabolite profiles indicative of mitochondrial abnormalities. Whole body protein turnover was unchanged. These findings demonstrate mitochondrial abnormalities concomitant with whole body and skeletal muscle functional derangements associated with human cancer, supporting future work studying the role of mitochondria in the muscle deficits associated with cancer.NEW & NOTEWORTHY To our knowledge, this is the first study to suggest that skeletal muscle mitochondrial deficits are associated with cancer-associated weight loss in humans. Mitochondrial deficits were concurrent with reductions in whole body and skeletal muscle functional capacity. Whether mitochondrial deficits are causal or secondary to cancer-associated weight loss and functional deficits remains to be determined, but this study supports further exploration of mitochondria as a driver of cancer-associated losses in muscle mass and function.

Concurrent Validity of the Garmin Vivofit®4 to Accurately Record Step Count in Older Adults in Challenging Environments

There is little evidence of the concurrent validity of commercially available wrist-worn long battery life activity monitors to measure steps in older adults at slow speeds and with real-world challenges. Forty adults aged over 60 years performed a treadmill protocol at four speeds, a 50-m indoor circuit, and a 200-m outdoor circuit with environmental challenges while wearing a Garmin Vivofit®4, the activPAL3™, and a chest-worn camera angled at the feet. The Garmin Vivofit®4 showed high intraclass correlation coefficients2,1 (.98-.99) and low absolute percentage error rates (<2%) at the fastest treadmill speeds and the outdoor circuit. Step counts were underestimated at the slowest treadmill speed and the indoor circuit. The Garmin Vivofit®4 is accurate for older adults at higher walking speeds and during outdoor walking. However, it underestimates steps at slow speeds and when walking indoors with postural transitions.

A Transparent Method for Step Detection using an Acceleration Threshold

Step-based metrics provide simple measures of ambulatory activity, yet device software either includes undisclosed proprietary step detection algorithms or simply do not compute step-based metrics. We aimed to develop and validate a simple algorithm to accurately detect steps across various ambulatory and non-ambulatory activities. Seventy-five adults (21-39 years) completed seven simulated activities of daily living (e.g., sitting, vacuuming, folding laundry) and an incremental treadmill protocol from 0.22-2.2ms^-1. Directly observed steps were hand-tallied. Participants wore GENEActiv and ActiGraph accelerometers, one of each on their waist and on their non-dominant wrist. Raw acceleration (g) signals from the anterior-posterior, medial-lateral, vertical, and vector magnitude (VM) directions were assessed separately for each device. Signals were demeaned across all activities and bandpass filtered [0.25, 2.5Hz]. Steps were detected via peak picking, with optimal thresholds (i.e., minimized absolute error from accumulated hand counted) determined by iterating minimum acceleration values to detect steps. Step counts were converted into cadence (steps/minute), and k-fold cross-validation quantified error (root mean squared error [RMSE]). We report optimal thresholds for use of either device on the waist (threshold=0.0267g) and wrist (threshold=0.0359g) using the VM signal. These thresholds yielded low error for the waist (RMSE<173 steps, ≤2.28 steps/minute) and wrist (RMSE<481 steps, ≤6.47 steps/minute) across all activities, and outperformed ActiLife's proprietary algorithm (RMSE=1312 and 2913 steps, 17.29 and 38.06 steps/minute for the waist and wrist, respectively). The thresholds reported herein provide a simple, transparent framework for step detection using accelerometers during treadmill ambulation and activities of daily living for waist- and wrist-worn locations.

Night-to-night variability of sleep apnea detected by cyclic variation of heart rate during long-term continuous ECG monitoring

Background

Sleep apnea is common in patients with cardiovascular disease and is a factor that worsens prognosis. Holter 24‐h ECG screening for sleep apnea is beneficial in the care of these patients, but due to high night‐to‐night variability of sleep apnea, it can lead to misdiagnosis and misclassification of disease severity.

Methods

To investigate the long‐term dynamic behavior of sleep apnea, seven‐day ECGs recorded with a patch ECG recorder in 120 patients were analyzed for the cyclic variation of heart rate (CVHR) during sleep periods as determined by a built‐in three‐axis accelerometer.

Results

The frequency of CVHR (Fcv) showed considerable night‐to‐night variability (coefficient of variance, 66 ± 35%), which was consistent with the night‐to‐night variability in apnea‐hypopnea index and oxygen desaturation index reported in earlier studies. In patients with presumed moderate‐to‐severe sleep apnea (Fcv > 15 cph at least one night), it was missed on 62% of nights, and on at least one night in 88% of patients. The CV of Fcv was negatively correlated with the average of Fcv, suggesting that patients with mild sleep apnea show greater night‐to‐night variability and would benefit from long‐term assessment. The average Fcv was higher in the supine position, but the night‐to‐night variability was not explained by the night‐to‐night variability of time spent in the supine position.

Conclusions

CVHR analysis of long‐term ambulatory ECG recordings is useful for improving the reliability of screening for sleep apnea without placing an extra burden on patients with cardiovascular disease and their care.

Accuracy and Diversity of Wearable Device-Based Gait Speed Measurement Among Older Men: Observational Study

Background

Gait speed measurements are widely used in clinical practice, as slow gait is a major predictor of frailty and a diagnostic criterion for sarcopenia. With the development of wearable devices, it is possible to estimate the gait speed in daily life by simply wearing the device.

Objective

This study aims to accurately determine the characteristics of daily life gait speed and analyze their association with sarcopenia.

Methods

We invited community-dwelling men aged >50 years who had visited the outpatient clinic at a tertiary university hospital to participate in the study. Daily life gait speed was assessed using a wearable smart belt (WELT) for a period of 4 weeks. Data from participants who wore the smart belt for at least 10 days during this period were included. After 4 weeks, data from a survey about medical and social history, usual gait speed measurements, handgrip strength measurements, and dual-energy x-ray absorptiometry were analyzed.

Results

A total of 217,578 daily life gait speed measurements from 106 participants (mean age 71.1, SD 7.6 years) were analyzed. The mean daily life gait speed was 1.23 (SD 0.26) m/s. The daily life gait speed of the participants varied according to the time of the day and day of the week. Daily life gait speed significantly slowed down with age (P<.001). Participants with sarcopenia had significantly lower mean daily life gait speed (mean 1.12, SD 0.11 m/s) than participants without sarcopenia (mean 1.23, SD 0.08 m/s; P<.001). Analysis of factors related to mean daily life gait speed showed that age and skeletal muscle mass of the lower limbs were significantly associated characteristics.

Conclusions

More diverse and accurate information about gait speed can be obtained by measuring daily life gait speed using a wearable device over an appropriate period, compared with one-time measurements performed in a laboratory setting. Importantly, in addition to age, daily life gait speed is significantly associated with skeletal muscle mass of the lower limbs.

Is the number of daily walking steps in sedentary workers affected by age, gender, body mass index, education, and overall energy expenditure?

BACKGROUND

Healthy adults should take 10,000 steps per day to gain the resulting health benefits. Knowledge regarding the individual characteristics associated with daily walking steps would enhance resource allocation to those most likely to benefit from the 10,000-steps-per-day campaign.

OBJECTIVE

To determine the extent to which age, gender, body mass index (BMI), education, and energy expenditure influence daily walking steps in white-collar workers and to assess the correlation of daily walking steps among pedometer, wristband activity tracker, and smartphone application.

METHODS

A cross-sectional study was conducted on 49 sedentary workers. Daily walking steps were simultaneously assessed by three activity trackers in free-living conditions for 7 consecutive days. Associations between daily walking steps and individual factors were examined using linear regression. Correlation tests were conducted to assess the association among the three devices.

RESULTS

Multiple regression analyses showed that BMI was associated with daily walking steps. A moderate to good correlation in daily walking steps was found between the wristband activity tracker and pedometer, as well as between the smartphone application and pedometer.

CONCLUSIONS

BMI influenced daily walking steps in white-collar workers. Daily walking steps assessed by the wristband activity tracker and smartphone application differed from those assessed by the pedometer.

The Effect of Sensor Placement and Number on Physical Activity Recognition and Energy Expenditure Estimation in Older Adults: Validation Study

BACKGROUND

Research has shown the feasibility of human activity recognition using wearable accelerometer devices. Different studies have used varying numbers and placements for data collection using sensors.

OBJECTIVE

This study aims to compare accuracy performance between multiple and variable placements of accelerometer devices in categorizing the type of physical activity and corresponding energy expenditure in older adults.

METHODS

In total, 93 participants (mean age 72.2 years, SD 7.1) completed a total of 32 activities of daily life in a laboratory setting. Activities were classified as sedentary versus nonsedentary, locomotion versus nonlocomotion, and lifestyle versus nonlifestyle activities (eg, leisure walk vs computer work). A portable metabolic unit was worn during each activity to measure metabolic equivalents (METs). Accelerometers were placed on 5 different body positions: wrist, hip, ankle, upper arm, and thigh. Accelerometer data from each body position and combinations of positions were used to develop random forest models to assess activity category recognition accuracy and MET estimation.

RESULTS

Model performance for both MET estimation and activity category recognition were strengthened with the use of additional accelerometer devices. However, a single accelerometer on the ankle, upper arm, hip, thigh, or wrist had only a 0.03-0.09 MET increase in prediction error compared with wearing all 5 devices. Balanced accuracy showed similar trends with slight decreases in balanced accuracy for the detection of locomotion (balanced accuracy decrease range 0-0.01), sedentary (balanced accuracy decrease range 0.05-0.13), and lifestyle activities (balanced accuracy decrease range 0.04-0.08) compared with all 5 placements. The accuracy of recognizing activity categories increased with additional placements (accuracy decrease range 0.15-0.29). Notably, the hip was the best single body position for MET estimation and activity category recognition.

CONCLUSIONS

Additional accelerometer devices slightly enhance activity recognition accuracy and MET estimation in older adults. However, given the extra burden of wearing additional devices, single accelerometers with appropriate placement appear to be sufficient for estimating energy expenditure and activity category recognition in older adults.

Movement at work: A comparison of real time location system, accelerometer and observational data from an office work environment

Office workers can spend significant periods of time being stationary whilst at work, with potentially serious health consequences. The development of effective health interventions could be aided by a greater understanding of the location and environmental context in which this stationary behaviour occurs. Real time location systems (RTLS) potentially offer the opportunity to gather this much needed information, but they have not been extensively trialled in office workplaces, nor rigorously compared against more familiar devices such as accelerometers. The aim of this paper was to determine whether an RTLS can measure and spatially locate the non-stationary and stationary behaviours of adults working in an office work environment. Data collected from a series of comparison studies undertaken in a commercial office building suggests that RTLS can measure the velocity at which people are moving and locate them, when stationary, with an accuracy of 0.668 m (SD 0.389). This opens up significant opportunities to further understand how people move within buildings, the indoor physical environmental influences on that movement, and the development of effective interventions to help people to move more whilst at work.

A review of the evolution of scientific literature on technology-assisted approaches using RGB-D sensors for musculoskeletal health monitoring

The human musculoskeletal (MSK) system (also known as the locomotor system) provides strength and assistance to perform functional tasks and daily life activities. The MSK health monitoring plays a vital role in maintaining the body mobility and quality of life. Manual approaches for musculoskeletal health monitoring are subjective and require a clinician's intervention. The evolution in motion tracking technology enables us to capture the fine details of body movements. The research community has proposed various approaches to help clinicians in diagnosis and monitor treatment sessions. This paper succinctly reviews the evolution of technology-assisted approaches for musculoskeletal health monitoring, using motion capture sensors. To streamline the search through the literature database, the PICOS framework and PRISMA method have been incorporated. The present study reviews methods to transform motion capture data into kinematics variables and factors that affect the tracking performance of RGB-D sensors. Furthermore, widely utilized time-series filters for skeletal data denoising and smoothing for kinematics analysis, stochastic models for movement modeling, rule-based and template-based approaches for rehabilitation exercises assessment, and telerehabilitation sessions for remote health monitoring are explored. This article analyzes skeletal tracking methods by providing advantages and drawbacks of the state of the art rehabilitation sessions assessment, skeletal joint kinematics analysis, and MSK Telerehabilitation approaches. It also discusses the possible future research avenues to improve musculoskeletal disorder diagnosis and treatment monitoring. Our review signifies that RGB-D sensor-based approaches are inexpensive and portable for disorder diagnosis and treatment monitoring. It can also be a viable option for clinicians to provide contactless healthcare access to patients in the current scenario of the COVID-19 pandemic.

Optimization and Validation of a Classification Algorithm for Assessment of Physical Activity in Hospitalized Patients

Low amounts of physical activity (PA) and prolonged periods of sedentary activity are common in hospitalized patients. Objective PA monitoring is needed to prevent the negative effects of inactivity, but a suitable algorithm is lacking. The aim of this study is to optimize and validate a classification algorithm that discriminates between sedentary, standing, and dynamic activities, and records postural transitions in hospitalized patients under free-living conditions. Optimization and validation in comparison to video analysis were performed in orthopedic and acutely hospitalized elderly patients with an accelerometer worn on the upper leg. Data segmentation window size (WS), amount of PA threshold (PA Th) and sensor orientation threshold (SO Th) were optimized in 25 patients, validation was performed in another 25. Sensitivity, specificity, accuracy, and (absolute) percentage error were used to assess the algorithm’s performance. Optimization resulted in the best performance with parameter settings: WS 4 s, PA Th 4.3 counts per second, SO Th 0.8 g. Validation showed that all activities were classified within acceptable limits (>80% sensitivity, specificity and accuracy, ±10% error), except for the classification of standing activity. As patients need to increase their PA and interrupt sedentary behavior, the algorithm is suitable for classifying PA in hospitalized patients.

Physical Activity Monitoring Using a Fitbit Device in Ischemic Stroke Patients: Prospective Cohort Feasibility Study

Background

Continuous tracking of ambulatory activity in real-world settings using step activity monitors has many potential uses. However, feasibility, accuracy, and correlation with performance measures in stroke patients have not been well-established.

Objective

The primary study objective was to determine adherence with wearing a consumer-grade step activity monitor, the Fitbit Charge HR, in home-going ischemic stroke patients during the first 90 days after hospital discharge. Secondary objectives were to (1) determine accuracy of step counts of the Fitbit Charge HR compared with a manual tally; (2) calculate correlations between the Fitbit step counts and the mobility performance scores at discharge and 30 days after stroke; (3) determine variability and change in weekly step counts over 90 days; and (4) evaluate patient experience with using the Fitbit Charge HR poststroke.

Methods

A total of 15 participants with recent mild ischemic stroke wore a Fitbit Charge HR for 90 days after discharge and completed 3 mobility performance tests from the National Institutes of Health Toolbox at discharge and Day 30: (1) Standing Balance Test, (2) 2-Minute Walk Endurance Test, and (3) 4-Meter Walk Gait Speed Test. Accuracy of step activity monitors was assessed by calculating differences in steps recorded on the step activity monitor and a manual tally during 2-minute walk tests.

Results

Participants had a mean age of 54 years and a median modified Rankin scale score of 1. Mean daily adherence with step activity monitor use was 83.6%. Mean daily step count in the first week after discharge was 4376. Daily step counts increased slightly during the first 30 days after discharge (average increase of 52.5 steps/day; 95% CI 32.2-71.8) and remained stable during the 30-90 day period after discharge. Mean step count difference between step activity monitor and manual tally was –4.8 steps (–1.8%). Intraclass correlation coefficients for step counts and 2-minute walk, standing balance, and 4-meter gait speed at discharge were 0.41 (95% CI –0.14 to 0.75), –0.12 (95% CI –0.67 to 0.64), and 0.17 (95% CI –0.46 to 0.66), respectively. Values were similarly poor at 30 days.

Conclusions

The use of consumer-grade Fitbit Charge HR in patients with recent mild stroke is feasible with reasonable adherence and accuracy. There was poor correlation between step counts and gait speed, balance, and endurance. Further research is needed to evaluate the association between step counts and other outcomes relevant to patients, including patient-reported outcomes and measures of physical function.

Step and Distance Measurement From a Low-Cost Consumer-Based Hip and Wrist Activity Monitor: Protocol for a Validity and Reliability Assessment

Background

Self-tracking via wearable and mobile technologies is becoming an essential part of personal health management. At this point, however, little information is available to substantiate the validity and reliability of low-cost consumer-based hip and wrist activity monitors, with regard more specifically to the measurements of step counts and distance traveled while walking.

Objective

The aim of our study is to assess the validity and reliability of step and distance measurement from a low-cost consumer-based hip and wrist activity monitor specific in various walking conditions that are commonly encountered in daily life. Specifically, this study is designed to evaluate whether and to what extent validity and reliability could depend on the sensor placement on the human body and the walking task being performed.

Methods

Thirty healthy participants will be instructed to wear four PBN 2433 (Nakosite) activity monitors simultaneously, with one placed on each hip and each wrist. Participants will attend two experimental sessions separated by 1 week. During each experimental session, two separate studies will be performed. In study 1, participants will be instructed to complete a 2-minute walk test along a 30-meter indoor corridor under 3 walking speeds: very slow, slow, and usual speed. In study 2, participants will be required to complete the following 3 conditions performed at usual walking speed: walking on flat ground, upstairs, and downstairs. Activity monitor measured step count and distance values will be computed along with the actual step count (determined from video recordings) and distance (measured using a measuring tape) to determine validity and reliability for each activity monitor placement and each walking condition.

Results

Participant recruitment and data collection began in January 2020. As of June 2020, we enrolled 8 participants. Dissemination of study results in peer-reviewed journals is expected in spring 2021.

Conclusions

To the best of our knowledge, this is the first study that examines the validity and reliability of step and distance measurement during walking using the PBN 2433 (Nakosite) activity monitor. Results of this study will provide beneficial information on the effects of activity monitor placement, walking speed, and walking tasks on the validity and reliability of step and distance measurement. We believe such information is of utmost importance to general consumers, clinicians, and researchers.

International Registered Report Identifier (IRRID)

DERR1-10.2196/21262

Assessing New Methods to Optimally Detect Episodes of Non-metabolic Heart Rate Variability Reduction as an Indicator of Psychological Stress in Everyday Life: A Thorough Evaluation of Six Methods

Frequent or chronic reduction in heart rate variability (HRV) is a powerful predictor of cardiovascular disease, and psychological stress has been suggested to be a co-determinant of this reduction. Recently, we evaluated various methods to measure additional HRV reduction in everyday life and to relate these reductions to psychological stress. In the current paper, we thoroughly evaluate these methods and add two new methods in both newly acquired and reanalyzed datasets. All of these methods use a subset of 24 h worth of HRV and movement data to do so: either the first 10 min of every hour, the full 24 h, a combination of 10 min from three consecutive hours, a classification of level of movement, the data from day n to detect episodes in day n + 1, or a range of activities during lab calibration. The method that used the full 24 h worth of data detected the largest percentage of episodes of reduced additional HRV that matched with self-reported stress levels, making this method the most promising, while using the first 10 min from three consecutive hours was a good runner-up.

Immersive Virtual Reality to Restore Natural Long-Range Autocorrelations in Parkinson's Disease Patients' Gait During Treadmill Walking

Effects of treadmill walking on Parkinson’s disease (PD) patients’ spatiotemporal gait parameters and stride duration variability, in terms of magnitude [coefficient of variation (CV)] and temporal organization [long range autocorrelations (LRA)], are known. Conversely, effects on PD gait of adding an optic flow during treadmill walking using a virtual reality headset, to get closer to an ecological walk, is unknown. This pilot study aimed to compare PD gait during three conditions: Overground Walking (OW), Treadmill Walking (TW), and immersive Virtual Reality on Treadmill Walking (iVRTW). Ten PD patients completed the three conditions at a comfortable speed. iVRTW consisted in walking at the same speed as TW while wearing a virtual reality headset reproducing an optic flow. Gait parameters assessed were: speed, step length, cadence, magnitude (CV) and temporal organization (evenly spaced averaged Detrended Fluctuation Analysis, α exponent) of stride duration variability. Motion sickness was assessed after TW and iVRTW using the Simulator Sickness Questionnaire (SSQ). Step length was greater (p = 0.008) and cadence lower (p = 0.009) during iVRTW compared to TW while CV was similar (p = 0.177). α exponent was similar during OW (0.77 ± 0.07) and iVRTW (0.76 ± 0.09) (p = 0.553). During TW, α exponent (0.85 ± 0.07) was higher than during OW (p = 0.039) and iVRTW (p = 0.016). SSQ was similar between TW and iVRTW (p = 0.809). iVRTW is tolerable, could optimize TW effects on spatiotemporal parameters while not increasing CV in PD. Furthermore, iVRTW could help to capture the natural LRA of PD gait in laboratory settings and could potentially be a challenging second step in PD gait rehabilitation.

Reliability and Validity of Commercially Available Wearable Devices for Measuring Steps, Energy Expenditure, and Heart Rate: Systematic Review

Background

Consumer-wearable activity trackers are small electronic devices that record fitness and health-related measures.

Objective

The purpose of this systematic review was to examine the validity and reliability of commercial wearables in measuring step count, heart rate, and energy expenditure.

Methods

We identified devices to be included in the review. Database searches were conducted in PubMed, Embase, and SPORTDiscus, and only articles published in the English language up to May 2019 were considered. Studies were excluded if they did not identify the device used and if they did not examine the validity or reliability of the device. Studies involving the general population and all special populations were included. We operationalized validity as criterion validity (as compared with other measures) and construct validity (degree to which the device is measuring what it claims). Reliability measures focused on intradevice and interdevice reliability.

Results

We included 158 publications examining nine different commercial wearable device brands. Fitbit was by far the most studied brand. In laboratory-based settings, Fitbit, Apple Watch, and Samsung appeared to measure steps accurately. Heart rate measurement was more variable, with Apple Watch and Garmin being the most accurate and Fitbit tending toward underestimation. For energy expenditure, no brand was accurate. We also examined validity between devices within a specific brand.

Conclusions

Commercial wearable devices are accurate for measuring steps and heart rate in laboratory-based settings, but this varies by the manufacturer and device type. Devices are constantly being upgraded and redesigned to new models, suggesting the need for more current reviews and research.

Accelerometer based assessment of daily physical activity and sedentary time in adolescents with idiopathic scoliosis

BACKGROUND

Studies have shown a positive correlation between higher physical activity (PA) and health benefits. However, device-based assessment of PA and sedentary time (ST) in people with adolescent idiopathic scoliosis (AIS) has not been deeply investigated.

OBJECTIVE

Analysis and comparison of weekend and weekdays PA and ST using multiple accelerometers in people with AIS with different curvature severity compared to healthy controls.

METHODS

24 participants with AIS divided into 2 groups of 12 with Cobb angles < 40° and > 40°, along with 12 age and BMI matched healthy controls. Daily PA and ST during four consecutive days were measured using four tri-axial accelerometers. Clinical functional assessment was performed using the scoliosis research society (SRS-22) questionnaire.

RESULTS

The combined weekend and weekdays average daily step count was found to be 22% and 29% lower in the AIS groups with Cobb angle < 40° and > 40°, respectively, compared to the controls. The average ST was also reported to be 5% and 7% higher in the AIS groups with Cobb angle < 40° and > 40°, respectively, compared to the controls. The reported differences were significant in the AIS group with higher Cobb angle (p≤0.05). No significant differences in PA or ST were reported between the AIS groups based on curvature severity.

CONCLUSIONS

Decreased PA and increased ST observed in patients with AIS may have long term health implications and may play a role in the disease process. The device-based assessment of PA to understand potential benefits in clinical practice is recommended.

The validity of activity trackers is affected by walking speed: the criterion validity of Garmin Vivosmart® HR and StepWatch™ 3 for measuring steps at various walking speeds under controlled conditions

Introduction

The use of activity trackers has increased both among private consumers and in healthcare. It is therefore relevant to consider whether a consumer-graded activity tracker is comparable to or may substitute a research-graded activity tracker, which could further increase the use of activity trackers in healthcare and rehabilitation. Such use will require knowledge of their accuracy as the clinical implications may be significant. Studies have indicated that activity trackers are not sufficiently accurate, especially at lower walking speeds. The present study seeks to inform decision makers and healthcare personnel considering implementing physical activity trackers in clinical practice. This study investigates the criterion validity of the consumer-graded Garmin Vivosmart^® HR and the research-graded StepWatch^™ 3 compared with manual step count (gold standard) at different walking speeds under controlled conditions.

Methods

Thirty participants, wearing Garmin Vivosmart^® HR at the wrist and StepWatch^™ 3 at the ankle, completed six trials on a treadmill at different walking speeds: 1.6 km/h, 2.4 km/h, 3.2 km/h, 4.0 km/h, 4.8 km/h, and 5.6 km/h. The participants were video recorded, and steps were registered by manual step count. Medians and inter-quartile ranges (IQR) were calculated for steps and differences in steps between manually counted steps and the two devices. In order to assess the clinical relevance of the tested devices, the mean absolute percentage error (MAPE) was determined at each speed. A MAPE ≤3% was considered to be clinically irrelevant. Furthermore, differences between manually counted steps and steps recorded by the two devices were presented in Bland-Altman style plots.

Results

The median of differences in steps between Garmin Vivosmart^® HR and manual step count ranged from -49.5 (IQR = 101) at 1.6 km/h to -1 (IQR = 4) at 4.0 km/h. The median of differences in steps between StepWatch^™ 3 and manual step count were 4 (IQR = 14) at 1.6 km/h and 0 (IQR = 1) at all other walking speeds. The results of the MAPE showed that differences in steps counted by Garmin Vivosmart^® HR were clinically irrelevant at walking speeds 3.2-4.8 km/h (MAPE: 0.61-1.27%) as the values were below 3%. Differences in steps counted by StepWatch^™ 3 were clinically irrelevant at walking speeds 2.4-5.6 km/h (MAPE: 0.08-0.35%).

Conclusion

Garmin Vivosmart^® HR tended to undercount steps compared with the manual step count, and StepWatch^™ 3 slightly overcounted steps compared with the manual step count. Both the consumer-graded activity tracker (Garmin Vivosmart^® HR) and the research-graded (StepWatch^™ 3) are valid in detecting steps at selected walking speeds in healthy adults under controlled conditions. However, both activity trackers miscount steps at slow walking speeds, and the consumer graded activity tracker also miscounts steps at fast walking speeds.

Differences in Physical Activity and Sedentary Behavior Patterns of Postmenopausal Women With Normal vs. Low Total Hip Bone Mineral Density

Purpose: Recent evidence suggests that sedentary behavior (SB) may be associated with bone health. This study compares free-living physical activity (PA) and SB distribution patterns of postmenopausal women with normal vs. low total hip bone mineral density (BMD). Methods: Sixty nine post-menopausal women [mean (min-max) age: 61 (46-79) years] wore ActiGraph GT3X+ activity monitors on the bilateral ankles for 7 days in free-living. Participants were split into two groups: those with normal hip BMD (T-scores ≥-1.0; N = 34) and those with low hip BMD (T-scores <-1.0; N = 35) as defined by the World Health Organization. Daily active time, step counts, sedentary time, sedentary break number, and median sedentary bout length were estimated from ankle acceleration data. The distribution and accumulation patterns of time spent in sedentary bouts, sedentary breaks, and stepping bouts, and sedentary break and stepping bout lengths' variability were also investigated. Group differences were assessed using two-sampled t-tests and Mann-Whitney U-tests with significance levels of 0.5. Results: Significant between group differences (p < 0.05) were in total daily active time [median (IQR): 257 (209-326) vs. 249 (199-299) min], step count [14,188 (10,938-18,646) vs. 13,204 (10,337-16,630) steps], sedentary time [669 (584-731) vs. 687 (615-753) min], and sedentary break number [93 (68-129) breaks vs. 88 (64-113) breaks], as well as median sedentary bout length [15.1 (11.9-22.1) vs. 15.8 (12.1-24.9) min]. Participants' sedentary bouts were found to be power law distributed with 52% of sedentary time occurring in bouts ≥20 min for the normal BMD group, and 58% for the low BMD group. Significant differences were observed between groups in sedentary bouts' and sedentary breaks' power distribution exponents (p < 0.0001) and patterns of sedentary and stepping time accumulation using the Gini index (p ≤ 0.0014). Variability was significantly lower for sedentary break and stepping bout lengths for the low BMD group (p ≤ 0.0001). Participants with lower hip BMD have longer sedentary bouts with shorter and less complex activity bouts compared to participants with normal hip BMD. Conclusion: The results suggest healthier hip BMD may be associated with PA distributed more evenly throughout the day with shorter sedentary bouts. PA distribution should be considered in exercise-based bone health management programs.

Wearable Inertial Sensors to Assess Gait during the 6-Minute Walk Test: A Systematic Review

Wearable sensors are becoming increasingly popular for complementing classical clinical assessments of gait deficits. The aim of this review is to examine the existing knowledge by systematically reviewing a large number of papers focusing on the use of wearable inertial sensors for the assessment of gait during the 6-minute walk test (6MWT), a widely recognized, simple, non-invasive, low-cost and reproducible exercise test. After a systematic search on PubMed and Scopus databases, two raters evaluated the quality of 28 full-text articles. Then, the available knowledge was summarized regarding study design, subjects enrolled (number of patients and pathological condition, if any, age, male/female ratio), sensor characteristics (type, number, sampling frequency, range) and body placement, 6MWT protocol and extracted parameters. Results were critically discussed to suggest future directions for the use of inertial sensor devices in the clinics.

Wearable Sensors Quantify Mobility in People With Lower Limb Amputation During Daily Life

It is necessary to effectively assess functional mobility for appropriate prosthetic prescription and post-amputation rehabilitation. As part of this process, patients' ability for variable cadence and community ambulation are assessed in-clinic, often through visual assessments and without objective standards. The purpose of this study was to explore the clinical viability of using wearable sensors to characterize the functional mobility of people with lower limb amputation. We collected inertial measurement unit (IMU) and global positioning system (GPS) data over two weeks, from 17 individuals with lower limb amputation and 14 healthy non-amputee controls. We calculated stride-by-stride cadence, walking speed and stride lengths, along with whether they occurred in or out of the home. Self-selected walking speed was also assessed in the lab. Compared to the lab, both groups walked slower and with a lower cadence during their daily lives. There were no differences in cadence variability between groups or between strides taken in and out of the home. Both groups walked faster and with greater stride lengths away from the homes. The results suggest that functional capacity measured in the lab was not necessarily reflected in routine walking during daily life. The walking measures derived in this approach can be used to aid in the prosthetic prescription process or in the assessment of different interventions.

Accuracy of Optical Heart Rate Sensing Technology in Wearable Fitness Trackers for Young and Older Adults: Validation and Comparison Study

BACKGROUND

Wearable fitness trackers are devices that can record and enhance physical activity among users. Recently, photoplethysmography (PPG) devices that use optical heart rate sensors to detect heart rate in real time have become popular and help in monitoring and controlling exercise intensity. Although the benefits of using optical heart rate monitors have been highlighted through studies, the accuracy of the readouts these commercial devices generate has not been widely assessed for different age groups, especially for the East Asian population with Fitzpatrick skin type III or IV.

OBJECTIVE

This study aimed to examine the accuracy of 2 wearable fitness trackers with PPG to monitor heart rate in real time during moderate exercise in young and older adults.

METHODS

A total of 20 young adults and 20 older adults were recruited for this study. All participants were asked to undergo a series of sedentary and moderate physical activities using indoor aerobic exercise equipment. In this study, the Polar H7 chest-strapped heart rate monitor was used as the criterion measure in 2 fitness trackers, namely Xiaomi Mi Band 2 and Garmin Vivosmart HR+. The real-time, second-by-second heart rate data obtained from both devices were recorded using the broadcast heart rate mode. To critically analyze the results, multiple statistical parameters including the mean absolute percentage error (MAPE), Lin concordance correlation coefficient (CCC), intraclass correlation coefficient, the Pearson product moment correlation coefficient, and the Bland-Altman coefficient were determined to examine the performances of the devices.

RESULTS

Both test devices exhibited acceptable overall accuracy as heart rate sensors based on several statistical tests. Notably, the MAPE values were below 10% (the designated threshold) in both devices (GarminYoung=3.77%; GarminSenior=4.73%; XiaomiYoung=7.69%; and XiaomiSenior=6.04%). The scores for reliability test of CCC for Garmin were 0.92 (Young) and 0.80 (Senior), whereas those for Xiaomi were 0.76 (Young) and 0.73 (Senior). However, the results obtained using the Bland-Altman analysis indicated that both test optical devices underestimated the average heart rate. More importantly, the study documented some unexpected outlier readings reported by these devices when used on certain participants.

CONCLUSIONS

The study reveals that commonly used optical heart rate sensors, such as the ones used herein, generally produce accurate heart rate readings irrespective of the age of the user. However, users should avoid relying entirely on these readings to indicate exercise intensities, as these devices have a tendency to produce erroneous, extreme readings, which might misinterpret the real-time exercise intensity. Future studies should therefore emphasize the occurrence rate of such errors, as this will likely benefit the development of improved models of heart rate sensors.

Design and Validation of a Minimal Complexity Algorithm for Stair Step Counting

Wearable sensors play a significant role for monitoring the functional ability of the elderly and in general, promoting active ageing. One of the relevant variables to be tracked is the number of stair steps (single stair steps) performed daily, which is more challenging than counting flight of stairs and detecting stair climbing. In this study, we proposed a minimal complexity algorithm composed of a hierarchical classifier and a linear model to estimate the number of stair steps performed during everyday activities. The algorithm was calibrated on accelerometer and barometer recordings measured using a sensor platform worn at the wrist from 20 healthy subjects. It was then tested on 10 older people, specifically enrolled for the study. The algorithm was then compared with other three state-of-the-art methods, which used the accelerometer, the barometer or both. The experiments showed the good performance of our algorithm (stair step counting error: 13.8%), comparable with the best state-of-the-art (p > 0.05), but using a lower computational load and model complexity. Finally, the algorithm was successfully implemented in a low-power smartwatch prototype with a memory footprint of about 4 kB.

Verification, analytical validation, and clinical validation (V3): the foundation of determining fit-for-purpose for Biometric Monitoring Technologies (BioMeTs)

Digital medicine is an interdisciplinary field, drawing together stakeholders with expertize in engineering, manufacturing, clinical science, data science, biostatistics, regulatory science, ethics, patient advocacy, and healthcare policy, to name a few. Although this diversity is undoubtedly valuable, it can lead to confusion regarding terminology and best practices. There are many instances, as we detail in this paper, where a single term is used by different groups to mean different things, as well as cases where multiple terms are used to describe essentially the same concept. Our intent is to clarify core terminology and best practices for the evaluation of Biometric Monitoring Technologies (BioMeTs), without unnecessarily introducing new terms. We focus on the evaluation of BioMeTs as fit-for-purpose for use in clinical trials. However, our intent is for this framework to be instructional to all users of digital measurement tools, regardless of setting or intended use. We propose and describe a three-component framework intended to provide a foundational evaluation framework for BioMeTs. This framework includes (1) verification, (2) analytical validation, and (3) clinical validation. We aim for this common vocabulary to enable more effective communication and collaboration, generate a common and meaningful evidence base for BioMeTs, and improve the accessibility of the digital medicine field.

A review of computational approaches for evaluation of rehabilitation exercises

Recent advances in data analytics and computer-aided diagnostics stimulate the vision of patient-centric precision healthcare, where treatment plans are customized based on the health records and needs of every patient. In physical rehabilitation, the progress in machine learning and the advent of affordable and reliable motion capture sensors have been conducive to the development of approaches for automated assessment of patient performance and progress toward functional recovery. The presented study reviews computational approaches for evaluating patient performance in rehabilitation programs using motion capture systems. Such approaches will play an important role in supplementing traditional rehabilitation assessment performed by trained clinicians, and in assisting patients participating in home-based rehabilitation. The reviewed computational methods for exercise evaluation are grouped into three main categories: discrete movement score, rule-based, and template-based approaches. The review places an emphasis on the application of machine learning methods for movement evaluation in rehabilitation. Related work in the literature on data representation, feature engineering, movement segmentation, and scoring functions is presented. The study also reviews existing sensors for capturing rehabilitation movements and provides an informative listing of pertinent benchmark datasets. The significance of this paper is in being the first to provide a comprehensive review of computational methods for evaluation of patient performance in rehabilitation programs.

Comparison of gait speeds from wearable camera and accelerometer in structured and semi-structured environments

A feasibility study was conducted to investigate the use of a wearable gait analysis system for classifying gait speed using a low-cost wearable camera in a semi-structured indoor setting. Data were collected from 19 participants who wore the system during indoor walk sequences at varying self-determined speeds (slow, medium, and fast). Gait parameters using this system were compared with parameters obtained from a vest comprising of a single triaxial accelerometer and from a marker-based optical motion-capture system. Computer-vision techniques and signal processing methods were used to generate frequency-domain gait parameters from each gait-recording device, and those parameters were analysed to determine the effectiveness of the different measurement systems in discriminating gait speed. Results indicate that the authors’ low-cost, portable, vision-based system can be effectively used for in-home gait analysis.

Use of accelerometer-based activity monitoring in orthopaedics: benefits, impact and practical considerations

Studies of the effectiveness of orthopaedic interventions do not generally measure physical activity (PA). Applying accelerometer-based activity monitoring in orthopaedic studies will add relevant information to the generally examined physical function and pain assessment.Accelerometer-based activity monitoring is practically feasible in orthopaedic patient populations, since current day activity sensors have battery time and memory to measure continuously for several weeks without requiring technical expertise.The ongoing development in sensor technology has made it possible to combine functional tests with activity monitoring.For clinicians, the application of accelerometer-based activity monitoring can provide a measure of PA and can be used for clinical comparisons before and after interventions.In orthopaedic rehabilitation, accelerometer-based activity monitoring may be used to help patients reach their targets for PA and to coach patients towards a more active lifestyle through direct feedback. Cite this article: EFORT Open Rev 2019;4:678-685. DOI: 10.1302/2058-5241.4.180041.

Quantitative assessment of lower limbs gross motor function in children with cerebral palsy based on surface EMG and inertial sensors

Taking advantage of motion sensing technology, a quantitative assessment method for lower limbs motor function of cerebral palsy (CP) based on the gross motor function measurement (GMFM)-24 scale was explored in this study. According to the motion analysis on GMFM-24 scale, we translated the assessment problem of GMFM-24 scale into a detection problem of different motion modes including static state, fall, step, turning, alternating gait, walking, running, lifting legs, kicking balls, and jumping. The surface electromyography (sEMG) electrodes and inertial sensors were adopted to capture motion data, and a framework integrating a series of detection algorithms was presented for the assessment of lower limbs gross motor function. Two groups of participants including 8 healthy adults and 14 CP children were recruited. A self-developed data acquisition equipment integrating 24 sEMG electrodes and 9 inertial units was adopted for data acquisition. A platform based on two laser beam sensors was used to perform cross-border detection. The parameters/thresholds of motion detection algorithms were determined by the data from healthy adults, and the lower limbs gross motor function evaluation was conducted on 14 CP children. The experimental results verified the feasibility and effectiveness of the proposed quantitative assessment method. Compared to the clinical assessment score based on GMFM-24 scale, 90.1% accuracy was obtained for evaluation of 303 tasks in 14 CP children. The objective motor function assessment method proposed has potential application value for the quantitative assessment of lower limbs motor function of CP children in clinical practice. Graphical abstract The algorithm framework for the assessment of lower limbs gross motor function. Using the GMFM-24 scale as the evaluation standard, a quantitative evaluation program for the lower limbs gross motor function of CP children based on motion sensing technology was proposed.

Outcome Measures of Free-Living Activity in Spinal Cord Injury Rehabilitation

PURPOSE OF REVIEW

The purpose of this article was to describe the utilization of body worn activity monitors in the SCI population and discuss the challenges of using body worn sensors in rehabilitation research.

RECENT FINDINGS

Many activity monitor-based measures have been used and validated in the SCI population including stroke number, push frequency, upper limb activity counts and wheelchair propulsion distance measured from a sensor attached to the wheelchair.

SUMMARY

The ability to accurately measure physical activity in the free-living environment using body-worn sensors has the potential to enhance the understanding of barriers to adequate activity and identify possible effective interventions. As the use of activity monitors used in SCI rehabilitation research continues to grow, care must be taken to overcome challenges related to participant adherence and data quality.

Validity of accelerometry in step detection and gait speed measurement in orthogeriatric patients

BACKGROUND

Mobile accelerometry is a powerful and promising option to capture long-term changes in gait in both clinical and real-world scenarios. Increasingly, gait parameters have demonstrated their value as clinical outcome parameters, but validation of these parameters in elderly patients is still limited.

OBJECTIVE

The aim of this study was to implement a validation framework appropriate for elderly patients and representative of real-world settings, and to use this framework to test and improve algorithms for mobile accelerometry data in an orthogeriatric population.

METHODS

Twenty elderly subjects wearing a 3D-accelerometer completed a parcours imitating a real-world scenario. High-definition video and mobile reference speed capture served to validate different algorithms.

RESULTS

Particularly at slow gait speeds, relevant improvements in accuracy have been achieved. Compared to the reference the deviation was less than 1% in step detection and less than 0.05 m/s in gait speed measurements, even for slow walking subjects (< 0.8 m/s).

CONCLUSION

With the described setup, algorithms for step and gait speed detection have successfully been validated in an elderly population and demonstrated to have improved performance versus previously published algorithms. These results are promising that long-term and/or real-world measurements are possible with an acceptable accuracy even in elderly frail patients with slow gait speeds.

Accelerometry data in health research: challenges and opportunities

Wearable accelerometers provide detailed, objective, and continuous measurements of physical activity (PA). Recent advances in technology and the decreasing cost of wearable devices led to an explosion in the popularity of wearable technology in health research. An ever-increasing number of studies collect high-throughput, sub-second level raw acceleration data. In this paper, we discuss problems related to the collection and analysis of raw accelerometry data and refer to published solutions. In particular, we describe the size and complexity of the data, the within- and between-subject variability, and the effects of sensor location on the body. We also discuss challenges related to sampling frequency, device calibration, data labeling and multiple PA monitors synchronization. We illustrate these points using the Developmental Epidemiological Cohort Study (DECOS), which collected raw accelerometry data on individuals both in a controlled and the free-living environment.

Concurrent validity of the Fitbit for assessing sedentary behavior and moderate-to-vigorous physical activity

Background

Recent advances in sensor technologies have promoted the use of consumer-based accelerometers such as Fitbit Flex in epidemiological and clinical research; however, the validity of the Fitbit Flex in measuring sedentary behavior (SED) and physical activity (PA) has not been fully determined against previously validated research-grade accelerometers such as ActiGraph GT3X+. Therefore, the purpose of this study was to examine the concurrent validity of the Fitbit Flex against ActiGraph GT3X+ in a free-living condition.

Methods

A total of 65 participants (age: M = 42, SD = 14 years, female: 72%) each wore a Fitbit Flex and GT3X+ for seven consecutive days. After excluding sleep and non-wear time, time spent (min/day) in SED and moderate-to-vigorous PA (MVPA) were estimated using various cut-points for GT3X+ and brand-specific algorithms for Fitbit, respectively. Repeated measures one-way ANOVA and mean absolute percent errors (MAPE) served to examine differences and measurement errors in SED and MVPA estimates between Fitbit Flex and GT3X+, respectively. Pearson and Spearman correlations and Bland-Altman (BA) plots were used to evaluate the association and potential systematic bias between Fitbit Flex and GT3X+. PROC MIXED procedure in SAS was used to examine the equivalence (i.e., the 90% confidence interval with ±10% equivalence zone) between the devices.

Results

Fitbit Flex produced similar SED and low MAPE (mean difference [MD] = 37 min/day, P = .21, MAPE = 6.8%), but significantly higher MVPA and relatively large MAPE (MD = 59–77 min/day, P < .0001, MAPE = 56.6–74.3%) compared with the estimates from GT3X+ using three different cut-points. The correlations between Fitbit Flex and GT3X+ were consistently higher for SED (r = 0.90, ρ = 0.86, P < .01), but weaker for MVPA (r = 0.65–0.76, ρ = 0.69–0.79, P < .01). BA plots revealed that there is no apparent bias in estimating SED.

Conclusion

In comparison with the GT3X+ accelerometer, the Fitbit Flex provided comparatively accurate estimates of SED, but the Fitbit Flex overestimated MVPA under free-living conditions. Future investigations using the Fitbit Flex should be aware of present findings.

Locomotion and cadence detection using a single trunk-fixed accelerometer: validity for children with cerebral palsy in daily life-like conditions

BACKGROUND

Physical therapy interventions for ambulatory youth with cerebral palsy (CP) often focus on activity-based strategies to promote functional mobility and participation in physical activity. The use of activity monitors validated for this population could help to design effective personalized interventions by providing reliable outcome measures. The objective of this study was to devise a single-sensor based algorithm for locomotion and cadence detection, robust to atypical gait patterns of children with CP in the real-life like monitoring conditions.

METHODS

Study included 15 children with CP, classified according to Gross Motor Function Classification System (GMFCS) between levels I and III, and 11 age-matched typically developing (TD). Six IMU devices were fixed on participant's trunk (chest and low back/L5), thighs, and shanks. IMUs on trunk were independently used for development of algorithm, whereas the ensemble of devices on lower limbs were used as reference system. Data was collected according to a semi-structured protocol, and included typical daily-life activities performed indoor and outdoor. The algorithm was based on detection of peaks associated to heel-strike events, identified from the norm of trunk acceleration signals, and included several processing stages such as peak enhancement and selection of the steps-related peaks using heuristic decision rules. Cadence was estimated using time- and frequency-domain approaches. Performance metrics were sensitivity, specificity, precision, error, intra-class correlation coefficient, and Bland-Altman analysis.

RESULTS

According to GMFCS, CP children were classified as GMFCS I (n = 7), GMFCS II (n = 3) and GMFCS III (n = 5). Mean values of sensitivity, specificity and precision for locomotion detection ranged between 0.93-0.98, 0.92-0.97 and 0.86-0.98 for TD, CP-GMFCS I and CP-GMFCS II-III groups, respectively. Mean values of absolute error for cadence estimation (steps/min) were similar for both methods, and ranged between 0.51-0.88, 1.18-1.33 and 1.94-2.3 for TD, CP-GMFCS I and CP-GMFCS II-III groups, respectively. The standard deviation was higher in CP-GMFCS II-III group, the lower performances being explained by the high variability of atypical gait patterns.

CONCLUSIONS

The algorithm demonstrated good performance when applied to a wide range of gait patterns, from normal to the pathological gait of highly affected children with CP using walking aids.

Effects of Gait Strategy and Speed on Regularity of Locomotion Assessed in Healthy Subjects Using a Multi-Sensor Method

The regularity of pseudo-periodic human movements, including locomotion, can be assessed by autocorrelation analysis of measurements using inertial sensors. Though sensors are generally placed on the trunk or pelvis, movement regularity can be assessed at any body location. Pathological factors are expected to reduce regularity either globally or on specific anatomical subparts. However, other non-pathological factors, including gait strategy (walking and running) and speed, modulate locomotion regularity, thus potentially confounding the identification of the pathological factor. The present study’s objectives were (1) to define a multi-sensor method based on the autocorrelation analysis of the acceleration module (norm of the acceleration vector) to quantify regularity; (2) to conduct an experimental study on healthy adult subjects to quantify the effect on movement regularity of gait strategy (walking and running at the same velocity), gait speed (four speeds, lower three for walking, upper two for running), and sensor location (on four different body parts). Twenty-five healthy adults participated and four triaxial accelerometers were located on the seventh cervical vertebra (C7), pelvis, wrist, and ankle. The results showed that increasing velocity was associated with increasing regularity only for walking, while no difference in regularity was observed between walking and running. Regularity was generally highest at C7 and ankle, and lowest at the wrist. These data confirm and complement previous literature on regularity assessed on the trunk, and will support future analyses on individuals or groups with specific pathologies affecting locomotor functions.

Design and validation of a simple automated optical step counting method for treadmill walking

BACKGROUND

Reliable step counting is a critical part of locomotion research. Current counting methods can be inaccurate, time consuming, expensive or encumbering to the subject. Here, we present a camera-based optical method for automatically counting steps.

METHODS

Fifteen healthy adults walked, jogged and ran on a treadmill at three different constant speeds (1.21, 2.01, 2.68 m/s) and once at varying speed (1.21-2.68 m/s) for 90 s. Subjects had visual marker affixed to their left foot while walking. Video was recorded synchronously at low- and high-resolution during trials. The step count found manually from the video was compared to an automated video analysis system using the two configurations of the optical system.

RESULTS

Bland-Altman plots, Intra-class correlation coefficients (ICC) and relative error comparison were used for quantitative assessment of device reliability. Reliability of optical method was high (ICC ≥0.98).

CONCLUSIONS

The method produces accurate step count results for the range of speeds tested. They use customisable open-source software and off-the-shelf hardware. The method has a low cost of implementation compared to many consumer products and grants researchers access to the raw sensor data.

Multi-Functional Soft Strain Sensors for Wearable Physiological Monitoring

Wearable devices which monitor physiological measurements are of significant research interest for a wide number of applications including medicine, entertainment, and wellness monitoring. However, many wearable sensing systems are highly rigid and thus restrict the movement of the wearer, and are not modular or customizable for a specific application. Typically, one sensor is designed to model one physiological indicator which is not a scalable approach. This work aims to address these limitations, by developing soft sensors and including conductive particles into a silicone matrix which allows sheets of soft strain sensors to be developed rapidly using a rapid manufacturing process. By varying the morphology of the sensor sheets and electrode placement the response can be varied. To demonstrate the versatility and range of sensitivity of this base sensing material, two wearable sensors have been developed which show the detection of different physiological parameters. These include a pressure-sensitive insole sensor which can detect ground reaction forces and a strain sensor which can be worn over clothes to allow the measurements of heart rate, breathing rate, and gait.

Wearable Hardware Design for the Internet of Medical Things (IoMT)

As the life expectancy of individuals increases with recent advancements in medicine and quality of living, it is important to monitor the health of patients and healthy individuals on a daily basis. This is not possible with the current health care system in North America, and thus there is a need for wireless devices that can be used from home. These devices are called biomedical wearables, and they have become popular in the last decade. There are several reasons for that, but the main ones are: expensive health care, longer wait times, and an increase in public awareness about improving quality of life. With this, it is vital for anyone working on wearables to have an overall understanding of how they function, how they were designed, their significance, and what factors were considered when the hardware was designed. Therefore, this study attempts to investigate the hardware components that are required to design wearable devices that are used in the emerging context of the Internet of Medical Things (IoMT). This means that they can be used, to an extent, for disease monitoring through biosignal capture. In particular, this review study covers the basic components that are required for the front-end of any biomedical wearable, and the limitations that these wearable devices have. Furthermore, there is a discussion of the opportunities that they create, and the direction that the wearable industry is heading in.

Functional assessment and satisfaction of transfemoral amputees with low mobility (FASTK2): A clinical trial of microprocessor-controlled vs. non-microprocessor-controlled knees

BACKGROUND

The benefits of a microprocessor-controlled knee are well documented in transfemoral amputees who are unlimited community ambulators. There have been suggestions that transfemoral amputees with limited community ambulation will also benefit from a microprocessor-controlled knee. Current medical policy restricts microprocessor-controlled knees to unlimited community ambulators and, thereby, potentially limits function. This clinical trial was performed to determine if limited community ambulators would benefit from a microprocessor-controlled knee.

METHODS

50 unilateral transfemoral amputees, mean age 69, were tested using their current non-microprocessor-controlled knee, fit with a microprocessor-controlled knee and allowed 10 weeks of acclimation before being tested, and then retested with their original mechanical knee after 4 weeks of re-acclimation. Patient function was assessed in the free-living environment using tri-axial accelerometers. Patient satisfaction and safety were also measured.

FINDINGS

The subjects demonstrated improved outcomes when using the microprocessor-controlled knee. Subjects had a significant reduction in falls, spent less time sitting, and increased their activity level. Subjects also reported significantly better ambulation, improved appearance, and greater utility.

INTERPRETATION

This clinical trial demonstrated that transfemoral amputees with limited mobility clearly benefit from a microprocessor-controlled knee. Notably, a reduction in falls occurred while the subjects engaged in more physical activity, which resulted in increased subject satisfaction. The increased activity resulted in a greater exposure to fall risk, but that risk was moderated by the advanced technology. ClinicalTrials.gov No: NCT02240186.