Association of Accelerometer-Measured Physical Activity and Sedentary Behavior With Incident Cardiovascular Disease, Myocardial Infarction, and Ischemic Stroke: The Women's Health Study

Background Few studies have investigated associations of acclerometer-based assessments of physical activity (PA) and sedentary behavior (SB) with incidence of cardiovascular disease (CVD) and its components. This prospective cohort study assessed the associations of accelerometer-measured PA and SB with total CVD, myocardial infarction, and ischemic stroke (IS). Methods and Results The authors included 16 031 women aged 62 years and older, free of CVD, with adherent accelerometer wear (≥10 hours/day for ≥4 days) from the Women's Health Study (mean age, 71.4 years [SD, 5.6 years]). Hip-worn ActiGraph GT3X+ accelerometers measured total volume of PA (total average daily vector magnitude), minutes per day of high-light PA and moderate to vigorous PA (MVPA), and SB. Women reported diagnoses of CVD, which were adjudicated using medical records and death certificates. Hazard ratios (HRs) were estimated for each exposure, and 95% CIs using Cox proportional hazards models were adjusted for accelerometer wear time, age, self-reported general health, postmenopausal hormone therapy, smoking status, and alcohol use. The hypothetical effect of replacing 10 minutes/day of SB or high-light PA with MVPA on CVD incidence was assessed using adjusted isotemporal substitution Cox models. Over a mean of 7.1 years (SD, 1.6 years) of follow-up, 482 total CVD cases, 107 myocardial infarction cases, and 181 IS cases were diagnosed. Compared with the lowest quartiles of total average daily vector magnitude and MVPA (≤60 minutes), women who were in the highest quartiles (>120 minutes of MVPA) had a 43% (95% CI, 24%-58%) and 38% (95% CI, 18%-54%) lower hazard of total CVD, respectively. Estimates were similar for total average daily vector magnitude and MVPA with IS, but PA was not associated with myocardial infarction overall. High-light PA was not associated with any CVD outcomes. Women who spent <7.4 hours sedentary per day had a 33% (95% CI, 11%-49%) lower hazard of total CVD compared with those who spent ≥9.5 hours sedentary. Replacing 10 minutes of SB with MVPA was associated with a 4% lower incidence of total CVD (HR, 0.96 [95% CI, 0.93-0.99]). Conclusions Accelerometer-assessed total PA and MVPA were inversely associated with total CVD and IS incidence, and SB was directly associated with total CVD; high-light PA was not related to CVD.

A Case for Unifying Accelerometry-Derived Movement Behaviors and Tests of Exercise Capacity for the Assessment of Relative Physical Activity Intensity

Albert Einstein taught us that "everything is relative." People's experience of physical activity (PA) is no different, with "relativism" particularly pertinent to the perception of intensity. Markers of absolute and relative intensities of PA have different but complimentary utilities, with absolute intensity considered best for PA guideline adherence and relative intensity for personalized exercise prescription. Under the paradigm of exercise and PA as medicine, our Technical Note proposes a method of synchronizing accelerometry with the incremental shuttle walking test to facilitate description of the intensity of the free-living PA profile in absolute and relative terms. Our approach is able to generate and distinguish "can do" or "cannot do" (based on exercise capacity) and "does do" or "does not do" (based on relative intensity PA) classifications in a chronic respiratory disease population, facilitating the selection of potential appropriate individually tailored interventions. By synchronizing direct assessments of exercise capacity and PA, clearer insights into the intensity of PA performed during everyday life can be gleaned. We believe the next steps are as follows: (1) to determine the feasibility and effectiveness of using relative and absolute intensities in combination to personalize the approach, (2) to determine its sensitivity to change following interventions (eg, exercise-based rehabilitation), and (3) to explore the use of this approach in healthier populations and in other long-term conditions.

Machine Learning on Prediction of Relative Physical Activity Intensity Using Medical Radar Sensor and 3D Accelerometer

BACKGROUND

One of the most critical topics in sports safety today is the reduction in injury risks through controlled fatigue using non-invasive athlete monitoring. Due to the risk of injuries, it is prohibited to use accelerometer-based smart trackers, activity measurement bracelets, and smart watches for recording health parameters during performance sports activities. This study analyzes the synergy feasibility of medical radar sensors and tri-axial acceleration sensor data to predict physical activity key performance indexes in performance sports by using machine learning (ML). The novelty of this method is that it uses a 24 GHz Doppler radar sensor to detect vital signs such as the heartbeat and breathing without touching the person and to predict the intensity of physical activity, combined with the acceleration data from 3D accelerometers.

METHODS

This study is based on the data collected from professional athletes and freely available datasets created for research purposes. A combination of sensor data management was used: a medical radar sensor with no-contact remote sensing to measure the heart rate (HR) and 3D acceleration to measure the velocity of the activity. Various advanced ML methods and models were employed on the top of sensors to analyze the vital parameters and predict the health activity key performance indexes. three-axial acceleration, heart rate data, age, as well as activity level variances.

RESULTS

The ML models recognized the physical activity intensity and estimated the energy expenditure on a realistic level. Leave-one-out (LOO) cross-validation (CV), as well as out-of-sample testing (OST) methods, have been used to evaluate the level of accuracy in activity intensity prediction. The energy expenditure prediction with three-axial accelerometer sensors by using linear regression provided 97-99% accuracy on selected sports (cycling, running, and soccer). The ML-based RPE results using medical radar sensors on a time-series heart rate (HR) dataset varied between 90 and 96% accuracy. The expected level of accuracy was examined with different models. The average accuracy for all the models (RPE and METs) and setups was higher than 90%.

CONCLUSIONS

The ML models that classify the rating of the perceived exertion and the metabolic equivalent of tasks perform consistently.

Validation of Oura ring energy expenditure and steps in laboratory and free-living

BACKGROUND

Commercial activity trackers are increasingly used in research and compared with research-based accelerometers are often less intrusive, cheaper, with improved storage and battery capacity, although typically less validated. The present study aimed to determine the validity of Oura Ring step-count and energy expenditure (EE) in both laboratory and free-living.

METHODS

Oura Ring EE was compared against indirect calorimetry in the laboratory, followed by a 14-day free-living study with 32 participants wearing an Oura Ring and reference monitors (three accelerometers positioned at hip, thigh, and wrist, and pedometer) to evaluate Oura EE variables and step count.

RESULTS

Strong correlations were shown for Oura versus indirect calorimetry in the laboratory (r = 0.93), and versus reference monitors for all variables in free-living (r ≥ 0.76). Significant (p < 0.05) mean differences for Oura versus reference methods were found for laboratory measured sitting (- 0.12 ± 0.28 MET), standing (- 0.27 ± 0.33 MET), fast walk (- 0.82 ± 1.92 MET) and very fast run (- 3.49 ± 3.94 MET), and for free-living step-count (2124 ± 4256 steps) and EE variables (MET: - 0.34-0.26; TEE: 362-494 kcal; AEE: - 487-259 kcal). In the laboratory, Oura tended to underestimate EE with increasing discrepancy as intensity increased. The combined activities and slow running in the laboratory, and all MET placements, TEE hip and wrist, and step count in free-living had acceptable measurement errors (< 10% MAPE), whereas the remaining free-living variables showed close to (≤13.2%) acceptable limits.

CONCLUSION

This is the first study investigating the validity of Oura Ring EE against gold standard methods. Oura successfully identified major changes between activities and/or intensities but was less responsive to detailed deviations within activities. In free-living, Oura step-count and EE variables tightly correlated with reference monitors, though with systemic over- or underestimations indicating somewhat low intra-individual validity of the ring versus the reference monitors. However, the correlations between the devices were high, suggesting that the Oura can detect differences at group-level for active and total energy expenditure, as well as step count.

How many days are needed? Measurement reliability of wearable device data to assess physical activity

INTRODUCTION/PURPOSE

Physical activity studies often utilize wearable devices to measure participants' habitual activity levels by averaging values across several valid observation days. These studies face competing demands-available resources and the burden to study participants must be balanced with the goal to obtain reliable measurements of a person's longer-term average. Information about the number of valid observation days required to reliably measure targeted metrics of habitual activity is required to inform study design.

METHODS

To date, the number of days required to achieve a desired level of aggregate long-term reliability (typically 0.80) has often been estimated by applying the Spearman-Brown Prophecy formula to short-term test-retest reliability data from studies with single, relatively brief observation windows. Our work, in contrast, utilizes a resampling-based approach to quantify the long-term test-retest reliability of aggregate measures of activity in a cohort of 79 participants who were asked to wear a FitBit Flex every day for approximately one year.

RESULTS

The conventional approach can produce reliability estimates that substantially overestimate the actual test-retest reliability. Six or more valid days of observation for each participant appear necessary to obtain 0.80 reliability for the average amount of time spent in light physical activity; 8 and 10 valid days are needed for sedentary time and moderate/vigorous activity respectively.

CONCLUSION

Protocols that result in 7-10 valid observation days for each participant may be needed to obtain reliable measurements of key physical activity metrics.

Validation of an Activity Type Recognition Model Classifying Daily Physical Behavior in Older Adults: The HAR70+ Model

Activity monitoring combined with machine learning (ML) methods can contribute to detailed knowledge about daily physical behavior in older adults. The current study (1) evaluated the performance of an existing activity type recognition ML model (HARTH), based on data from healthy young adults, for classifying daily physical behavior in fit-to-frail older adults, (2) compared the performance with a ML model (HAR70+) that included training data from older adults, and (3) evaluated the ML models on older adults with and without walking aids. Eighteen older adults aged 70-95 years who ranged widely in physical function, including usage of walking aids, were equipped with a chest-mounted camera and two accelerometers during a semi-structured free-living protocol. Labeled accelerometer data from video analysis was used as ground truth for the classification of walking, standing, sitting, and lying identified by the ML models. Overall accuracy was high for both the HARTH model (91%) and the HAR70+ model (94%). The performance was lower for those using walking aids in both models, however, the overall accuracy improved from 87% to 93% in the HAR70+ model. The validated HAR70+ model contributes to more accurate classification of daily physical behavior in older adults that is essential for future research.

Generalizability and performance of methods to detect non-wear with free-living accelerometer recordings

Wearable physical activity sensors are widely used in research and practice as they provide objective measures of human behavior at a low cost. An important challenge for accurate assessment of physical activity behavior in free-living is the detection non-wear. Traditionally, heuristic algorithms that rely on specific interval lengths have been employed to detect non-wear time; however, machine learned models are emerging. We explore the potential of detecting non-wear using decision trees that combine raw acceleration and skin temperature, and we investigate the generalizability of our models, traditional heuristic algorithms, and recently developed machine learned models by external validation. The Decision tree models were trained using one week of data from thigh- and hip-worn accelerometers from 64 children. External validation was performed using data from wrist-worn accelerometers of 42 adolescents. For non-wear episodes longer than 60 min, the heuristic algorithms performed the best with F1-scores above 0.96. However, regarding episodes shorter than 60 min, the best performing method was the decision tree model including the six most important predictors with F1 scores above 0.74 for all sensor locations. We conclude that for classifying non-wear time, researchers should carefully select an appropriate method and we encourage the use of external validation when reporting on machine learned non-wear models.

Real-Time Risk Assessment Detection for Weak People by Parallel Training Logical Execution of a Supervised Learning System Based on an IoT Wearable MEMS Accelerometer

Activity monitoring has become a necessary demand for weak people to guarantee their safety. The paper proposed a Parallel Training Logical Execution (PTLE) system using machine learning (ML) models on a microelectromechanical system (MEMS) accelerometer to detect coughs, falls, and other normal activities. When there are many categories, the ML prediction can be confused between these activities with each other. The PTLE system trains several models in parallel with more specific activity classes in each dataset. The shared tasks between parallel models relieve the complexity for a single one. There are six additional parameters for accelerometer characteristics, which were calculated from three axes accelerations as input features to improve the ML's consciousness. Once all models were trained, the system was ready to receive the input accelerations and activated the logical flow to manage link operation between these ML models for output predictions. Random Forest (RF) had the highest potential among the ML classification algorithms after the validation. In the experiment, the comparison between the PTLE model and the regular ML model were carried out with real-time data from an M5stickC wearable device on the user's chest to the trained models on PC. The result showed the advancement of the proposed method in term of precision, recall, F1-score with an overall accuracy of 98% in the real-time test. The accelerations from the wearable device were sent to ML models via Wi-Fi with Message Queue Telemetry Transport (MQTT) broker, and the activity predictions were transferred to the cloud for the family members or doctor care based on Internet of Things (IoT) communication.

Development and Evaluation of Sedentary Time Cut-Points for the activPAL in Adults Using the GGIR R-Package

The purpose of this study was to develop sedentary cut-points for the activPAL and evaluate their performance against a criterion measure (i.e., activPAL processed by PALbatch). Part 1: Thirty-five adults (23.4 ± 3.6 years) completed 12 laboratory activities (6 sedentary and 6 non-sedentary activities). Receiver operator characteristic (ROC) curves proposed optimal Euclidean Norm Minus One (ENMO) and Mean Amplitude Deviation (MAD) cut-points of 26.4 mg (ENMO) and 30.1 mg (MAD). Part 2: Thirty-eight adults (22.6 ± 4.1 years) wore an activPAL during free-living. Estimates from PALbatch and MAD revealed a mean percent error (MPE) of 2.2%, mean absolute percent error (MAPE) of 6.5%, limits of agreement (LoA) of 19% with absolute and relative equivalence zones of 5% and 0.3 SD. Estimates from PALbatch and ENMO revealed an MPE of -10.6%, MAPE of 14.4%, LoA of 31% and 16% and 1 SD equivalence zones. After standing was isolated from sedentary behaviours, ROC analysis proposed an optimal cut-off of 21.9 mg (herein ENMOs). Estimates from PALbatch and ENMOs revealed an MPE of 3.1%, MAPE of 7.5%, LoA of 25% and 9% and 0.5 SD equivalence zones. The MAD and ENMOs cut-points performed best in discriminating between sedentary and non-sedentary activity during free-living.

Effect of Equipment on the Accuracy of Accelerometer-Based Human Activity Recognition in Extreme Environments

A little explored area of human activity recognition (HAR) is in people operating in relation to extreme environments, e.g., mountaineers. In these contexts, the ability to accurately identify activities, alongside other data streams, has the potential to prevent death and serious negative health events to the operators. This study aimed to address this user group and investigate factors associated with the placement, number, and combination of accelerometer sensors. Eight participants (age = 25.0 ± 7 years) wore 17 accelerometers simultaneously during lab-based simulated mountaineering activities, under a range of equipment and loading conditions. Initially, a selection of machine learning techniques was tested. Secondly, a comprehensive analysis of all possible combinations of the 17 accelerometers was performed to identify the optimum number of sensors, and their respective body locations. Finally, the impact of activity-specific equipment on the classifier accuracy was explored. The results demonstrated that the support vector machine (SVM) provided the most accurate classifications of the five machine learning algorithms tested. It was found that two sensors provided the optimum balance between complexity, performance, and user compliance. Sensors located on the hip and right tibia produced the most accurate classification of the simulated activities (96.29%). A significant effect associated with the use of mountaineering boots and a 12 kg rucksack was established.

Accelerometer-Measured Physical Activity and Sedentary Time among Children in Japan before and during COVID-19: A Cross-Sectional and Longitudinal Analysis

This study examined changes in physical activity (PA), sedentary behavior (SB), screen time, sleep, and executive function among Japanese preschoolers between COVID-19 pre-pandemic and pandemic periods, using cross-sectional and longitudinal data. Accelerometer data from 63 children aged 5-6 years were collected from three kindergartens in Tokyo, Japan, in late 2019 (pre-COVID-19). This was compared to the data of 49 children aged 5-6 years from the same kindergartens, collected in late 2020 (during COVID-19). Sixteen children in the pre-COVID-19 cohort also participated in the 2020 survey and provided data for the longitudinal analysis. The mean minutes of PA, SB, screen time, and sleep duration, as well as executive function, were compared between the pre- and during COVID-19 cohorts. After adjusting for school, sex, and accelerometer wear time, there were no significant differences in any of the measured outcomes between the two cohorts. However, the analysis of longitudinal data revealed significant increases in time spent in SB and on screens, and a decrease in light-intensity PA and sleep duration during the pandemic compared to the pre-pandemic period. Results suggest that, despite the COVID-19 pandemic, young children's activity levels and SB did not significantly differ from pre-pandemic levels. However, school-aged children's SB, light PA, and sleep time were affected, although this cannot be disentangled from the effects of the transition to school.

Isotemporal substitution of sleep or sedentary behavior with physical activity in the context of frailty among older adults: a cross-sectional study

BACKGROUND

Frailty syndrome is associated with various physical, cognitive, social, economic, and environmental factors. Although frailty syndrome occurs progressively with age, prevention and treatment are possible. Reducing or eliminating risks and increasing protective factors may be potential strategies for reducing the prevalence of injuries related to frailty. One of the most effective actions is to decrease the time spent in sedentary behavior (SB) by increasing regular physical activity (PA).

OBJECTIVE

To examine the hypothetical effect of substitution of the time spent in sleep or SB with an equivalent time spent performing moderate or vigorous PA on frailty syndrome in the older population.

DESIGN AND SETTING

An analytical cross-sectional study conducted using exploratory methods of survey, carried out in Alcobaça city, Bahia, Brazil.

METHODS

A total of 456 older adults of both sexes, aged ≥ 60 years, participated in this study. Frailty syndrome was identified according to the criteria of the Study of Osteoporotic Fractures. PA and SB were assessed using the International Physical Activity Questionnaire, and sleep was assessed using the Pittsburgh Sleep Quality Index. The effects of time substitution on these behaviors were verified using Poisson regression.

RESULTS

The replacement of 60 min/day of SB (prevalence ratio, PR = 0.52; 95% confidence interval, CI: 0.28-0.96) or sleep (PR = 0.52; 95% CI: 0.27-0.98) with 60 min/day of moderate PA (MPA) was associated with a 48% reduction in the prevalence of frailty syndrome.

CONCLUSIONS

Replacing the time spent sitting or sleeping with the same amount of MPA time may reduce frailty; the longer the duration of time spent in the substitution of sleep or SB with MPA, the greater the benefits.

Accelerometer-Measured Latent Physical Activity Profiles and Neurocognition Among Middle-Aged and Older Hispanic/Latino Adults in the Hispanic Community Health Study/Study of Latinos (HCHS/SOL)

OBJECTIVES

Derive latent profiles of accelerometry-measured moderate-vigorous physical activity (MVPA) for Hispanic/Latino adults, examine associations between latent MVPA profiles and neurocognition, and describe profiles via self-reported MVPA.

METHODS

Complex survey design methods were applied to cross-sectional data from 7,672 adults ages 45-74 years in the Hispanic Community Health Study/Study of Latinos (HCHS/SOL; 2008-2011). MVPA was measured via hip-worn accelerometers. Latent profile analysis was applied to derive latent MVPA profiles (minutes/day of week). Neurocognition was assessed with the Brief-Spanish English Verbal Learning Test (B-SEVLT) Sum, B-SEVLT Recall, Controlled Oral Word Association Test (word fluency), and Digit Symbol Substitution (DSS) test. All tests were z-scored, and a global neurocognition score was generated by averaging across scores. Survey linear regression models were used to examine associations between latent MVPA profiles and neurocognitive measures. Self-reported MVPA domains were estimated (occupational, transportation, and recreational) for each latent profile.

RESULTS

Four latent MVPA profiles from the overall adult target population (18-74 years) were derived and putatively labeled: No MVPA, low, moderate, and high. Only the high MVPA profile (compared to moderate) was associated with lower global neurocognition. Sensitivity analyses using latent MVPA profiles with only participants aged 45-74 years showed similar profiles, but no associations between latent MVPA profiles and neurocognition. The occupational MVPA domain led in all latent MVPA profiles.

DISCUSSION

We found no consistent evidence to link accelerometry-measured MVPA profiles to neurocognitive function. Research to better characterize the role of high occupational MVPA in relation to neurocognition among Hispanic/Latino adults are needed.

Is Wrist Accelerometry Suitable for Threshold Scoring? A Comparison of Hip-Worn and Wrist-Worn ActiGraph Data in Low-Active Older Adults With Obesity

BACKGROUND

Hip- and wrist-worn ActiGraph accelerometers are widely used in research on physical activity as they offer an objective assessment of movement intensity across the day. Herein we characterize and contrast key structured physical activities and common activities of daily living via accelerometry data collected at the hip and wrist from a sample of community-dwelling older adults.

METHODS

Low-active, older adults with obesity (age 60+ years) were fit with an ActiGraph GT3X+ accelerometer on their nondominant wrist and hip before completing a series of tasks in a randomized order, including sitting/standing, sweeping, folding laundry, stair climbing, ambulation at different intensities, and cycling at different intensities. Participants returned a week later and completed the tasks once again. Vector magnitude counts/second were time-matched during each task and then summarized into counts/minute (CPM).

RESULTS

Monitors at both wear locations similarly characterized standing, sitting, and ambulatory tasks. A key finding was that light home chores (sweeping, folding laundry) produced higher and more variable CPM values than fast walking via wrist ActiGraph. Regression analyses revealed wrist CPM values were poor predictors of hip CPM values, with devices aligning best during fast walking (R2 = 0.25) and stair climbing (R2 = 0.35).

CONCLUSIONS

As older adults spend a considerable portion of their day in nonexercise activities of daily living, researchers should be cautious in the use of simply acceleration thresholds for scoring wrist-worn accelerometer data. Methods for better classifying wrist-worn activity monitor data in older adults are needed.

Accelerometers in Our Pocket: Does Smartphone Accelerometer Technology Provide Accurate Data?

This study evaluates accelerometer performance of three new state of the art smartphones and focuses on accuracy. The motivating research question was whether accelerator accuracy obtained with these off-the-shelf modern smartphone accelerometers was or was not statistically different from that of a gold-standard reference system. We predicted that the accuracy of the three modern smartphone accelerometers in human movement data acquisition do not differ from that of the Vicon MX motion capture system. To test this prediction, we investigated the comparative performance of three different commercially available current generation smartphone accelerometers among themselves and to a gold-standard Vicon MX motion capture system. A single subject design was implemented for this study. Pearson's correlation coefficients^® were calculated to verify the validity of the smartphones' accelerometer data against that of the Vicon MX motion capture system. The Intraclass Correlation Coefficient (ICC) was used to assess the smartphones' accelerometer performance reliability compared to that of the Vicon MX motion capture system. Results demonstrated that (a) the tested smartphone accelerometers are valid and reliable devices for estimating accelerations and (b) there were not significant differences among the three current generation smartphones and the Vicon MX motion capture system's mean acceleration data. This evidence indicates how well recent generation smartphone accelerometer sensors are capable of measuring human body motion. This study, which bridges a significant information gap between the accuracy of accelerometers measured close to production and their accuracy in actual smartphone research, should be interpreted within the confines of its scope, limitations and strengths. Further research is warranted to validate our arguments, suggestions, and results, since this is the first study on this topic.

Stepping towards More Intuitive Physical Activity Metrics with Wrist-Worn Accelerometry: Validity of an Open-Source Step-Count Algorithm

Stepping-based targets such as the number of steps per day provide an intuitive and commonly used method of prescribing and self-monitoring physical activity goals. Physical activity surveillance is increasingly being obtained from wrist-worn accelerometers. However, the ability to derive stepping-based metrics from this wear location still lacks validation and open-source methods. This study aimed to assess the concurrent validity of two versions (1. original and 2. optimized) of the Verisense step-count algorithm at estimating step-counts from wrist-worn accelerometry, compared with steps from the thigh-worn activPAL as the comparator. Participants (n = 713), across three datasets, had >24 h continuous concurrent accelerometry wear on the non-dominant wrist and thigh. Compared with activPAL, total daily steps were overestimated by 913 ± 141 (mean bias ± 95% limits of agreement) and 742 ± 150 steps/day with Verisense algorithms 1 and 2, respectively, but moderate-to-vigorous physical activity (MVPA) steps were underestimated by 2207 ± 145 and 1204 ± 103 steps/day in Verisense algorithms 1 and 2, respectively. In summary, the optimized Verisense algorithm was more accurate in detecting total and MVPA steps. Findings highlight the importance of assessing algorithm performance beyond total step count, as not all steps are equal. The optimized Verisense open-source algorithm presents acceptable accuracy for derivation of stepping-based metrics from wrist-worn accelerometry.

Machine Learning Models for Weight-Bearing Activity Type Recognition Based on Accelerometry in Postmenopausal Women

Hip-worn triaxial accelerometers are widely used to assess physical activity in terms of energy expenditure. Methods for classification in terms of different types of activity of relevance to the skeleton in populations at risk of osteoporosis are not currently available. This publication aims to assess the accuracy of four machine learning models on binary (standing and walking) and tertiary (standing, walking, and jogging) classification tasks in postmenopausal women. Eighty women performed a shuttle test on an indoor track, of which thirty performed the same test on an indoor treadmill. The raw accelerometer data were pre-processed, converted into eighteen different features and then combined into nine unique feature sets. The four machine learning models were evaluated using three different validation methods. Using the leave-one-out validation method, the highest average accuracy for the binary classification model, 99.61%, was produced by a k-NN Manhattan classifier using a basic statistical feature set. For the tertiary classification model, the highest average accuracy, 94.04%, was produced by a k-NN Manhattan classifier using a feature set that included all 18 features. The methods and classifiers within this study can be applied to accelerometer data to more accurately characterize weight-bearing activity which are important to skeletal health.

Detection of Parkinson's Disease Using Wrist Accelerometer Data and Passive Monitoring

Parkinson's disease is a neurodegenerative disorder impacting patients' movement, causing a variety of movement abnormalities. It has been the focus of research studies for early detection based on wearable technologies. The benefit of wearable technologies in the domain rises by continuous monitoring of this population's movement patterns over time. The ubiquity of wrist-worn accelerometry and the fact that the wrist is the most common and acceptable body location to wear the accelerometer for continuous monitoring suggests that wrist-worn accelerometers are the best choice for early detection of the disease and also tracking the severity of it over time. In this study, we use a dataset consisting of one-week wrist-worn accelerometry data collected from individuals with Parkinson's disease and healthy elderlies for early detection of the disease. Two feature engineering methods, including epoch-based statistical feature engineering and the document-of-words method, were used. Using various machine learning classifiers, the impact of different windowing strategies, using the document-of-words method versus the statistical method, and the amount of data in terms of number of days were investigated. Based on our results, PD was detected with the highest average accuracy value (85% ± 15%) across 100 runs of SVM classifier using a set of features containing features from every and all windowing strategies. We also found that the document-of-words method significantly improves the classification performance compared to the statistical feature engineering model. Although the best performance of the classification task between PD and healthy elderlies was obtained using seven days of data collection, the results indicated that with three days of data collection, we can reach a classification performance that is not significantly different from a model built using seven days of data collection.

Accelerometer Cut-Points for Physical Activity Assessment in Adults with Mild to Moderate Huntington's Disease: A Cross-Sectional Multicentre Study

Accelerometers can estimate the intensity, frequency, and duration of physical activity in healthy adults. Although thresholds to distinguish varying levels of activity intensity using the Actigraph wGT3X-B have been established for the general population, their accuracy for Huntington's disease (HD) is unknown. We aimed to define and cross-validate accelerometer cut-points for different walking speeds in adults with mild to moderate HD. A cross-sectional, multicentre, case-control, observational study was conducted with a convenience sample of 13 symptomatic ambulatory HD participants. The accelerometer was placed around the right hip, and a heart monitor was fitted around the chest to monitor heart rate variability. Participants walked on a treadmill at three speeds with light, moderate and vigorous intensities. Correlation and receiver operation curve analyses were performed between the accelerometer magnitude vector with relative oxygen and heart rate. Optimal cut-points for walking speeds of 3.2 km/h were ≤2852; 5.2 km/h: >2852 to ≤4117, and in increments until their maximum velocity: >4117. Our results support the application of the disease-specific cut-points for quantifying physical activity in patients with mild to moderate HD and promoting healthy lifestyle interventions.

Significant Features for Human Activity Recognition Using Tri-Axial Accelerometers

Activity recognition using wearable sensors has become essential for a variety of applications. Tri-axial accelerometers are the most widely used sensor for activity recognition. Although various features have been used to capture patterns and classify the accelerometer signals to recognise activities, there is no consensus on the best features to choose. Reducing the number of features can reduce the computational cost and complexity and enhance the performance of the classifiers. This paper identifies the signal features that have significant discriminative power between different human activities. It also investigates the effect of sensor placement location, the sampling frequency, and activity complexity on the selected features. A comprehensive list of 193 signal features has been extracted from accelerometer signals of four publicly available datasets, including features that have never been used before for activity recognition. Feature significance was measured using the Joint Mutual Information Maximisation (JMIM) method. Common significant features among all the datasets were identified. The results show that the sensor placement location does not significantly affect recognition performance, nor does it affect the significant sub-set of features. The results also showed that with high sampling frequency, features related to signal repeatability and regularity show high discriminative power.

Physical Activity in Centenarians beyond Cut-Point-Based Accelerometer Metrics

This study described and compared physical activity (PA) characteristics at the end of the human lifespan using conventional cut-point-based versus cut-point-free accelerometer metrics. Eighteen institutionalized centenarians (101.5 ± 2.1 years, 72.2% female, 89% frail) wore the wrist GENEActiv accelerometer for 7 days. Conventional metrics, such as time spent in light-intensity PA (LiPA) and moderate-to-vigorous intensity PA (MVPA) were calculated according to published cut-points for adults and older adults. The following cut-point-free metrics were evaluated: average acceleration, intensity gradient and Mx metrics. Depending on the cut-point, centenarians accumulated a median of 15-132 min/day of LiPA and 3-15 min/day of MVPA. The average acceleration was 9.2 mg [Q1: 6.7 mg-Q3: 12.6 mg] and the intensity gradient was -3.19 [-3.34--3.12]. The distribution of Z-values revealed positive skew for MVPA, indicating a potential floor effect, whereas the skew magnitude was attenuated for cut-point-free metrics such as intensity gradient or M5. However, both cut-point-based and cut-point-free metrics were similarly positively associated with functional independence, cognitive and physical capacities. This is the first time that PA has been described in centenarians using cut-point-free metrics. Our results suggest that new analytical approaches could overcome cut-point limitations when studying the oldest-old. Future studies using these new cut-point-free PA metrics are warranted to provide more complete and comparable information across groups and populations.

The influence of removing the 10-minute bout requirement on the demographic, behaviour and health profiles of Canadian adults who meet the physical activity recommendations

BACKGROUND

Recently, the Canadian 24-Hour Movement Guidelines for Adults were released, and included a revised physical activity (PA) recommendation. The recommendation of 150 minutes per week of moderate-to-vigorous intensity PA (MVPA) was revised, from requiring that MVPA be accrued in bouts of 10 minutes or more (bouted) to having no bout requirement (non-bouted). The objective of this study was to assess whether there were differences in sociodemographic, health and fitness characteristics of Canadians who met the bouted and non-bouted PA recommendations.

DATA AND METHODS

Using adult (aged 18 to 79 years) accelerometer data from three combined cycles of the nationally representative Canadian Health Measures Survey (N = 7,102), this study compared adherence to the bouted and non-bouted recommendations. Differences in sociodemographic, health and fitness measures were assessed using independent t-tests and chi-squares. Multivariate linear and logistic regressions controlling for age, sex, household education and smoking examined associations with health and fitness measures.

RESULTS

More adults met the PA recommendation using the non-bouted versus bouted (45.3% vs. 18.5%) requirement. Characteristics of those who met the bouted and only the non-bouted recommendations were similar. Exceptions among those who met only the non-bouted recommendation compared with meeting the bouted recommendation included fewer adults aged 65 years and older; lower MVPA, recreation PA and transport PA; and higher sedentary time, light PA and grip strength.

INTERPRETATION

Although the removal of the 10-minute bout requirement increased the proportion of Canadian adults who met the PA recommendation, there were no substantial differences in the sociodemographic and health characteristics of the populations captured by the bouted and non-bouted definitions. Results help to inform the transition in reporting for PA surveillance.

Investigating Wrist-Based Acceleration Summary Measures across Different Sample Rates towards 24-Hour Physical Activity and Sleep Profile Assessment

Wrist-worn wearable sensors have attracted considerable research interest because of their potential in providing continuous, longitudinal, non-invasive measurements, leading to insights into Physical Activity (PA), sleep, and circadian variability. Three key practical considerations for research-grade wearables are as follows: (a) choosing an appropriate sample rate, (b) summarizing raw three-dimensional accelerometry data for further processing (accelerometry summary measures), and (c) accurately estimating PA levels and sleep towards understanding participants' 24-hour profiles. We used the CAPTURE-24 dataset, where 148 participants concurrently wore a wrist-worn three-dimensional accelerometer and a wearable camera over approximately 24 h to obtain minute-by-minute labels: sleep; and sedentary light, moderate, and vigorous PA. We propose a new acceleration summary measure, the Rate of Change Acceleration Movement (ROCAM), and compare its performance against three established approaches summarizing three-dimensional acceleration data towards replicating the minute-by-minute labels. Moreover, we compare findings where the acceleration data was sampled at 10, 25, 50, and 100 Hz. We demonstrate the competitive advantage of ROCAM towards estimating the five labels (80.2% accuracy) and building 24-hour profiles where the sample rate of 10 Hz is fully sufficient. Collectively, these findings provide insights facilitating the deployment of large-scale longitudinal actigraphy data processing towards 24-hour PA and sleep-profile assessment.

On the Problem of State Recognition in Injection Molding Based on Accelerometer Data Sets

The last few decades have been characterised by a very active application of smart technologies in various fields of industry. This paper deals with industrial activities, such as injection molding, where it is required to monitor continuously the manufacturing process to identify both the effective running time and down-time periods. Supervised machine learning algorithms are developed to recognize automatically the periods of the injection molding machines. The former algorithm uses directly the features of the descriptive statistics, while the latter one utilizes a convolutional neural network. The automatic state recognition system is equipped with an 3D-accelerometer sensor whose datasets are used to train and verify the proposed algorithms. The novelty of our contribution is that accelerometer data-based machine learning models are used to distinguish producing and non-producing periods by means of recognition of key steps in an injection molding cycle. The first testing results show the approximate overall balanced accuracy of 72-92% that illustrates the large potential of the monitoring system with the accelerometer. According to the ANOVA test, there are no sufficient statistical differences between the comparative algorithms, but the results of the neural network exhibit higher variances of the defined accuracy metrics.

High Precision Classification of Resting and Eating Behaviors of Cattle by Using a Collar-Fitted Triaxial Accelerometer Sensor

Cattle are less active than humans. Hence, it was hypothesized in this study that transmitting acceleration signals at a 1 min sampling interval to reduce storage load has the potential to improve the performance of motion sensors without affecting the precision of behavior classification. The behavior classification performance in terms of precision, sensitivity, and the F1-score of the 1 min serial datasets segmented in 3, 4, and 5 min window sizes based on nine algorithms were determined. The collar-fitted triaxial accelerometer sensor was attached on the right side of the neck of the two fattening Korean steers (age: 20 months) and the steers were observed for 6 h on day one, 10 h on day two, and 7 h on day three. The acceleration signals and visual observations were time synchronized and analyzed based on the objectives. The resting behavior was most correctly classified using the combination of a 4 min window size and the long short-term memory (LSTM) algorithm which resulted in 89% high precision, 81% high sensitivity, and 85% high F1-score. High classification performance (79% precision, 88% sensitivity, and 83% F1-score) was also obtained in classifying the eating behavior using the same classification method (4 min window size and an LSTM algorithm). The most poorly classified behavior was the active behavior. This study showed that the collar-fitted triaxial sensor measuring 1 min serial signals could be used as a tool for detecting the resting and eating behaviors of cattle in high precision by segmenting the acceleration signals in a 4 min window size and by using the LSTM classification algorithm.