Predicting Energy Expenditure in Preschool Children Using Accelerometer and Gyroscope Data.

Estimating energy expenditure (EE) in real-world settings is crucial for studying human behavior and energy balance. Despite advances in wrist-worn inertial measurement units (IMU), actigraphy remains the most accepted measure for estimating EE, despite known Errors in accuracy, particularly in people with obesity. We developed an algorithm estimating EE from commercial smartwatch sensor data, and validated it against actigraphy-based energy estimates in people with obesity. In an in-lab study, 27 participants wore a Fossil Sport smartwatch and ActiGraph wGT3X+ while performing activities of varying intensities. Another 25 participants wore the smartwatch for 2 days in a free-living study. We built a machine learning model to estimate metabolic equivalent of task (MET) values/minute using smartwatch accelerometer and gyroscope data. Analysis included 2,189 minutes of in-lab and 14,045 minutes of free-living data. Compared to the metabolic cart, our model achieved lower root mean square error (0.28–0.32) across various sliding windows. In the free-living study, our algorithm’s estimates fell within SD of the best actigraphy-based estimates for 95.03% of minutes. Our proposed method accurately estimated METs compared to 11 algorithms primarily validated in non-obese populations, suggesting that commercial wrist-worn devices can provide more inclusive and reliable EE measures using our algorithm.

The aim of this study was to compare the energy expenditure (EE) estimations of activity-specific prediction equations (ASPE) and of an artificial neural network (ANNEE) based on accelerometry with measured EE. Forty-three children (age: 9.8 ± 2.4 yr) performed eight different activities. They were equipped with one tri-axial accelerometer that collected data in 1-s epochs and a portable gas analyzer. The ASPE and the ANNEE were trained to estimate the EE by including accelerometry, age, gender, and weight of the participants. To provide the activity-specific information, a decision tree was trained to recognize the type of activity through accelerometer data. The ASPE were applied to the activity-type-specific data recognized by the tree (Tree-ASPE). The Tree-ASPE precisely estimated the EE of all activities except cycling [bias: -1.13 ± 1.33 metabolic equivalent (MET)] and walking (bias: 0.29 ± 0.64 MET; P < 0.05). The ANNEE overestimated the EE of stationary activities (bias: 0.31 ± 0.47 MET) and walking (bias: 0.61 ± 0.72 MET) and underestimated the EE of cycling (bias: -0.90 ± 1.18 MET; P < 0.05). Biases of EE in stationary activities (ANNEE: 0.31 ± 0.47 MET, Tree-ASPE: 0.08 ± 0.21 MET) and walking (ANNEE 0.61 ± 0.72 MET, Tree-ASPE: 0.29 ± 0.64 MET) were significantly smaller in the Tree-ASPE than in the ANNEE (P < 0.05). The Tree-ASPE was more precise in estimating the EE than the ANNEE. The use of activity-type-specific information for subsequent EE prediction equations might be a promising approach for future studies.

ABSTRACTThis study compared accuracy of energy expenditure (EE) prediction models from accelerometer data collected in structured and simulated free-living settings. Twenty-four adults (mean age 45.8 years, 50% female) performed two sessions of 11 to 21 activities, wearing four ActiGraph GT9X Link activity monitors (right hip, ankle, both wrists) and a metabolic analyzer (EE criterion). Visit 1 (V1) involved structured, 5-min activities dictated by researchers; Visit 2 (V2) allowed participants activity choice and duration (simulated free-living). EE prediction models were developed incorporating data from one setting (V1/V2; V2/V2) or both settings (V1V2/V2). The V1V2/V2 method had the lowest root mean square error (RMSE) for EE prediction (1.04–1.23 vs. 1.10–1.34 METs for V1/V2, V2/V2), and the ankle-worn accelerometer had the lowest RMSE of all accelerometers (1.04–1.18 vs. 1.17–1.34 METs for other placements). The ankle-worn accelerometer and associated EE prediction models developed using data from both structured and simulated free-living settings should be considered for optimal EE prediction accuracy.

In studies that involve the use of a room calorimeter, 24 h energy intake is often larger than 24 h energy expenditure (24 h EE) because of a decrease in activity energy expenditure due to the confined space. This positive energy balance can have large consequences for the interpretation of substrate balances. The objective of this study was to develop a method for predicting an individual's 24 h EE in a room calorimeter at both low (1.4xRMR) and high (1.8xRMR) levels of physical activity. Two methods are presented that predict an individual's 24 h EE in a metabolic chamber. The first method was based on three components: (1) a 30 min measurement of resting metabolic rate (RMR) using a ventilated hood system; (2) measurement of exercise energy expenditure during 10 min of treadmill walking; and (3) estimation of free-living energy expenditure using a tri-axial motion sensor. Using these measurements we calculated the amount of treadmill time needed for each individual in order to obtain a total 24 h EE at either a low (1.4xRMR) or a high (1.8xRMR) level of physical activity. We also developed a method to predict total 24 h EE during the chamber stay by using the energy expenditure values for the different levels of activity as measured during the hours already spent in the chamber. This would provide us with a tool to adjust the exercise time and/or energy intake during the chamber stay. Method 1: there was no significant difference in expected and measured 24 h EE under either low (9.35+/-0.56 vs 9.51+/-0.47 MJ/day; measured vs predicted) or high activity conditions (13.41+/-0.74 vs 13.97+/-0.78 MJ/day; measured vs predicted). Method 2: the developed algorithm predicted 24 h EE for 97.6+/-4.0% of the final value at 3 h into the test day, and for 98.6+/-3.7% at 7 h into the test day. Both methods provide accurate prediction of energy expenditure in a room calorimeter at both high and low levels of physical activity. It equally shows that it is possible to accurately predict total 24 h EE from energy expenditure values obtained at 3 and 7 h into the study.

A thorough analysis of sports is becoming increasingly important during the training process of badminton players at both the recreational and professional level. Nowadays, game situations are usually filmed and reviewed afterwards in order to analyze the game situation, but these video set-ups tend to be difficult to analyze, expensive, and intrusive to set up. In contrast, we classified badminton movements using off-the-shelf accelerometer and gyroscope data. To this end, we organized a data capturing campaign and designed a novel neural network using different frame sizes as input. This paper shows that with only accelerometer data, our novel convolutional neural network is able to distinguish nine activities with 86% precision when using a sampling frequency of 50 Hz. Adding the gyroscope data causes an increase of up to 99% precision, as compared to, respectively, 79% and 88% when using a traditional convolutional neural network. In addition, our paper analyses the impact of different sensor placement options and discusses the impact of different sampling frequenciess of the sensors. As such, our approach provides a low cost solution that is easy to use and can collect useful information for the analysis of a badminton game.

This study explores Human Activity Recognition (HAR) using smartphone sensors to address the challenges posed by position-dependent datasets. We propose a position-independent system that leverages data from accelerometers, gyroscopes, linear accelerometers, and gravity sensors collected from smartphones placed either on the chest or in the left/right leg pocket. The performance of traditional machine learning algorithms (Decision Trees (DT), K-Nearest Neighbors (KNN), Random Forest (RF), Support Vector Classifier (SVC), and XGBoost) is compared against deep learning models (Gated Recurrent Unit (GRU), Long Short-Term Memory (LSTM), Temporal Convolutional Networks (TCN), and Transformer models) under two sensor configurations. Our findings highlight that the Temporal Convolutional Network (TCN) model consistently outperforms other models, particularly in the four-sensor non-overlapping configuration, achieving the highest accuracy of 97.70%. Deep learning models such as LSTM, GRU, and Transformer also demonstrate strong performance, showcasing their effectiveness in capturing temporal dependencies in HAR tasks. Traditional machine learning models, including RF and XGBoost, provide reasonable performance but do not match the accuracy of deep learning models. Additionally, incorporating data from linear accelerometers and gravity sensors led to slight improvements over using accelerometer and gyroscope data alone. This research enhances the recognition of passenger behaviors for intelligent transportation systems, contributing to more efficient congestion management and emergency response strategies.

A Validation of Equations for Predicting the Energy Expenditures of Hospitalized Patients

Falls are a major health risk that diminish the quality of life among elderly people. Apart from falls themselves, most dramatic consequences are usually related with long lying periods that can cause serious side effects. These findings call for pervasive long-term fall detection systems able to automatically detect falls. In this paper, we propose an effective fall detection algorithm for mobile platforms. Using data retrieved from wearable sensors, such as Inertial Measurements Units (IMUs) and/or SmartPhones (SPs), our algorithm is able to detect falls using features extracted from accelerometer and gyroscope. While mostly of the mobile-based solutions for fall management deal only with accelerometer data, in the proposed approach we combine the instantaneous acceleration magnitude vector with changes of the user's heading in a Threshold Based Algorithm (TBA). In such a way, we were able to handle falls detection with minimal computational load, increasing the overall system accuracy with respect to traditional fall management methods. Experimental results show the strong detection performance of the proposed solution in discriminating between falls and typical Activities of Daily Living (ADLs) presenting fall-like acceleration patterns.

Wearable monitors continue to play a critical role in scientific assessments of physical activity. Recently, research-grade monitors have begun providing raw data from photoplethysmography (PPG) alongside standard raw data from inertial sensors (accelerometers and gyroscopes). Raw PPG enables granular and transparent estimation of cardiovascular parameters such as heart rate, thus presenting a valuable alternative to standard PPG methodologies (most of which rely on consumer-grade monitors that provide only coarse output from proprietary algorithms). The implications for physical activity assessment are tremendous, since it is now feasible to monitor granular and concurrent trends in both movement and cardiovascular physiology using a single noninvasive device. However, new users must also be aware of challenges and limitations that accompany the use of raw PPG data. This viewpoint paper therefore orients new users to the opportunities and challenges of raw PPG data by presenting its mechanics, pitfalls, and availability, as well as its parallels and synergies with inertial sensors. This includes discussion of specific applications to the prediction of energy expenditure, activity type, and 24-hour movement behaviors, with an emphasis on areas in which raw PPG data may help resolve known issues with inertial sensing (eg, measurement during cycling activities). We also discuss how the impact of raw PPG data can be maximized through the use of open-source tools when developing and disseminating new methods, similar to current standards for raw accelerometer and gyroscope data. Collectively, our comments show the strong potential of raw PPG data to enhance the use of research-grade wearable activity monitors in science over the coming years.

1720 PURPOSE: The ability to accurately predict energy expenditure is essential for exercise prescription; however, little research is available when comparing the measured energy expenditure with prediction equations. Furthermore, little research is available comparing walking and running energy expenditure on a track to treadmill. The primary purpose of this study was to establish how well the published prediction equations for the energy expenditure of walking and running compare to the actual measured values. To accomplish this we reexamined whether differences exist in energy expenditure of walking vs. running 1600 meters, and if energy expenditure differences occur due to being on the track or treadmill which could confound the estimation of energy expenditure. METHODS: Indirect calorimetry was measured on 12 male and 12 female subjects while walking (4.8 km/hr) and running (9.6 km/hr) 1600 meters on the treadmill. A subgroup of 7 males and 10 females also performed the 1600 meter run/walk on the track. The measured energy expenditures were compared to prediction equations available in the literature. RESULTS: Running utilized significantly more energy for 1600 meters than walking in both genders (treadmill: running 481 ± 20, walking 340 ± 19 kJ; track: running 480 ± 23, walking 430 ± 14 kJ, P<0.01), and greater energy expenditures were found in the males than females (P<0.05). When adjusted for lean body mass, the greater energy expenditures for 1600 meters in males disappeared. There were no differences between the track and treadmill energy expenditures regardless of speed. The predictions using the ACSM or Léger equations for running did not differ significantly from the actual energy expenditures (total error-ACSM: 20 and −14.4 kJ for running and walking respectively; Légers walking: 10.1 kJ). Both the McArdle's table value for walking and running and van der Walt's prediction for walking significantly overestimated (P<0.01) energy expenditure for 1600 meters. CONCLUSIONS: ACSM's prediction model for walking and running as well as the Léger equation for walking appear to be the most suitable for the prediction of energy expenditure for both running and walking, while McArdle's table and van der Walt equation were not as strong predictors of energy expenditure.

Objectives: This study aimed to provide an improved energy expenditure estimation for heavy-load physical labor using accelerometer data and heart rate (HR) measured by wearables and to support food preparation and supply management for disaster relief and rescue operations as an expedition team. Methods: To achieve an individually optimized estimation for energy expenditure, a model equation parameter was determined based on the measurements of physical activity and HR during simulated rescue operations. The metabolic equivalent of task (MET), which was measured by using a tri-axial accelerometer and individual HR, was used, where two (minimum and maximum) or three (minimum, intermediate, and maximum) representative reference points were selected for each individual model fitting. In demonstrating the applicability of our approach in a realistic situation, accelerometer-based METs and HR of 30 males were measured using the tri-axial accelerometer and wearable HR during simulated rescue operations over 2days. Results: Data sets of 27 rescue operations (age:34.2 ± 7.5years; body mass index (BMI):22.9 ± 1.5kg/m2) were used for the energy expenditure estimation after excluding three rescue workers due to their activity type and insufficient HR measurement. Using the combined approach with a tri-axial accelerometer and HR, the total energy expenditure increased by 143% for two points and 133% for three points, compared with the estimated total energy expenditure using only the accelerometer-based method. Conclusion: The use of wearables provided a reasonable estimation of energy expenditure for physical workers with heavy equipment. The application of our approach to disaster relief and rescue operations can provide important insights into nutrition and healthcare management.

In the cloud age, it is quite easy to collect sensory data from smartphones. With these sensory data, it is desired to provide various kinds of applications to serve the user. In this research, we aim at developing an indoor navigation system on smartphone using solely smartphone sensory data. There are many researches on indoor localization and navigation in the literature. Nevertheless, environmental sensors and/or wearable sensors are usually needed. This can be costly and inconvenient. In this paper, we propose a smartphone indoor localization system using only accelerometer and gyroscope data from the smartphone. The Pedestrian Dead Reckoning (PDR) approach is used to build this system. The PDR approach is simple and efficient though seems traditional. The major weakness of the PDR is that the estimation error would accumulate over time. Thus we propose to add so-called calibration marks which look like short arrows and are placed on both the floor plan and the ground. To use the system, the user first finds a calibration mark on the ground, stands on it and faces the right direction. He/she then moves the android icon (representing the user) on top of the calibration mark on the floor plan on the smartphone. When the user starts to move, the android icon also moves on the floor plan following the real-time estimation of step length and moving direction change for each step from accelerometer and gyroscope data. This is a prototype of an indoor navigation system that can become fully functional after an optimal path planning module is included. Experimental results of estimated walking trace tests show high accuracy. The system is promising and useful as long as a floor plan and calibration marks are built in advance.

Validity of accelerometry for predicting physical activity and sedentary time in ambulatory children and young adults with cerebral palsy

BackgroundRecent advancements in wearable sensor technology have shown the feasibility of remote physical therapy at home. In particular, the current COVID-19 pandemic has revealed the need and opportunity of internet-based wearable technology in future health care systems. Previous research has shown the feasibility of human activity recognition technologies for monitoring rehabilitation activities in home environments; however, few comprehensive studies ranging from development to clinical evaluation exist.ObjectiveThis study aimed to (1) develop a home-based rehabilitation (HBR) system that can recognize and record the type and frequency of rehabilitation exercises conducted by the user using a smartwatch and smartphone app equipped with a machine learning (ML) algorithm and (2) evaluate the efficacy of the home-based rehabilitation system through a prospective comparative study with chronic stroke survivors.MethodsThe HBR system involves an off-the-shelf smartwatch, a smartphone, and custom-developed apps. A convolutional neural network was used to train the ML algorithm for detecting home exercises. To determine the most accurate way for detecting the type of home exercise, we compared accuracy results with the data sets of personal or total data and accelerometer, gyroscope, or accelerometer combined with gyroscope data. From March 2018 to February 2019, we conducted a clinical study with two groups of stroke survivors. In total, 17 and 6 participants were enrolled for statistical analysis in the HBR group and control group, respectively. To measure clinical outcomes, we performed the Wolf Motor Function Test (WMFT), Fugl-Meyer Assessment of Upper Extremity, grip power test, Beck Depression Inventory, and range of motion (ROM) assessment of the shoulder joint at 0, 6, and 12 months, and at a follow-up assessment 6 weeks after retrieving the HBR system.ResultsThe ML model created with personal data involving accelerometer combined with gyroscope data (5590/5601, 99.80%) was the most accurate compared with accelerometer (5496/5601, 98.13%) or gyroscope data (5381/5601, 96.07%). In the comparative study, the drop-out rates in the control and HBR groups were 40% (4/10) and 22% (5/22) at 12 weeks and 100% (10/10) and 45% (10/22) at 18 weeks, respectively. The HBR group (n=17) showed a significant improvement in the mean WMFT score (P=.02) and ROM of flexion (P=.004) and internal rotation (P=.001). The control group (n=6) showed a significant change only in shoulder internal rotation (P=.03).ConclusionsThis study found that a home care system using a commercial smartwatch and ML model can facilitate participation in home training and improve the functional score of the WMFT and shoulder ROM of flexion and internal rotation in the treatment of patients with chronic stroke. This strategy can possibly be a cost-effective tool for the home care treatment of stroke survivors in the future.Trial RegistrationClinical Research Information Service KCT0004818; https://tinyurl.com/y92w978t

a) Compare the predictive potential of speed and CSA(hip) (Computer Science Applications accelerometer positioned on the hip) for level terrain walking METs (1 MET = VO2 of 3.5 mL.kg(-1).min(-1)) and energy expenditure (kcal.min(-1)); b) cross-validate previously published CSA(hip)- and speed-based MET and energy expenditure prediction equations; c) measure self-paced walking speed, exercise intensity (METs) and energy expenditure in the middle aged population. Seventy-two 35- to 45-yr-old volunteers walked around a level, paved quadrangle at what they perceived to be a moderate pace. Oxygen consumption was measured using the criterion Douglas bag technique. Speed, CSA(hip), heart rate, and Borg rating of perceived exertion were also monitored. Speed explained 10% more variance of walking METs than CSA(hip). Speed and mass explained 8% more variance of walking energy expenditure (kcal.min) than CSA(hip) and mass. The best previously published regression equations predict our walking METs and energy expenditures within 95% prediction limits of +/- 0.7 METs and +/- 1.0 kcal.min(-1), respectively. Women paced themselves at a significantly higher mean speed (5.5 km.h(-1)) and intensity (4.1 METs) than their male counterparts (5.2 km.h(-1) and 3.8 METs). Both genders expended approximately 0.75 kcal.kg(-1) for every kilometer of level terrain walked. Speed-based MET and energy expenditure predictions during level terrain walking were more accurate than those utilizing CSA(hip).