Accurate Estimation Of Energy Expenditure Research Articles

Energy Expenditure Prediction from Accelerometry Data Using Long Short-Term Memory Recurrent Neural Networks.

The accurate estimation of energy expenditure from simple objective accelerometry measurements provides a valuable method for investigating the effect of physical activity (PA) interventions or population surveillance. Methods have been evaluated previously, but none utilize the temporal aspects of the accelerometry data. In this study, we investigated the energy expenditure prediction from acceleration measured at the subjects' hip, wrist, thigh, and back using recurrent neural networks utilizing temporal elements of the data. The acceleration was measured in children (N = 33) performing a standardized activity protocol in their natural environment. The energy expenditure was modelled using Multiple Linear Regression (MLR), stacked long short-term memory (LSTM) networks, and combined convolutional neural networks (CNN) and LSTM. The correlation and mean absolute percentage error (MAPE) were 0.76 and 19.9% for the MLR, 0.882 and 0.879 and 14.22% for the LSTM, and, with the combined LSTM-CNN, the best performance of 0.883 and 13.9% was achieved. The prediction error for vigorous intensities was significantly different (p < 0.01) from those of the other intensity domains: sedentary, light, and moderate. Utilizing the temporal elements of movement significantly improves energy expenditure prediction accuracy compared to other conventional approaches, but the prediction error for vigorous intensities requires further investigation.

Energy expenditure in healthy pregnant women: a systematic review.

Since pregnancy is a decisive period for women and children in terms of future health, it is very important to determine changes in energy expenditure and requirements. However, there is no reference method to determine the energy expenditure as there are many factors that change maternal energy homeostasis during this period. This review aims to evaluate the changes in energy expenditure during pregnancy and the effects of these changes on maternal weight gain, as well as the methods used to determine energy expenditure. Google Scholar, PubMed, ScienceDirect, Cochrane Library, and Wiley Online Library databases were searched to determine studies on energy expenditure during pregnancy between 2010 and 2022. Of the 2095 articles initially identified, 20 met the inclusion criteria for this review. Two independent reviewers selected and extracted data from those articles. The articles reviewed show that prepregnancy body mass index, gestational weight gain, changes in fat mass, and lean body mass influence energy expenditure during pregnancy. Doubly-labeled water and indirect calorimetric methods are considered reference methods for determining energy expenditure in pregnancy. Indirect calorimetric methods, when used together with body weight and physical activity measurements, allow an accurate estimation of resting energy expenditure and energy requirements for pregnant women. Metabolic Holter or accelerometers are valid methods for determining activity energy expenditure in pregnant women and are suitable for use during activity. Resting energy expenditure and activity energy expenditure should be carefully monitored and measured by valid methods to achieve the correct energy intake targets during pregnancy. The existing equations used in determining energy expenditure should be developed for a more accurate estimation of energy expenditure during pregnancy, or new pregnancy-specific equations should be derived.

Exercise Intensity and Activity Energy Expenditure of Professional Golf Players in Official Competitive Setting.

Research regarding the physical needs of professional golf players is lacking. With advances in wearable technology, it has become easier to analyze physiological responses such as heart rate (HR) to determine activity energy expenditure (AEE). The purpose of the study was to evaluate exercise intensity (EI) and AEE during 4 consecutive tournament's golf rounds using a popular wrist-based HR monitoring. Wearable systems for HR monitoring can be used to provide an accurate estimate of energy expenditure. Cross-sectional study. Level 3. A total of 20 male professional golfers participated in the study. Each player was monitored during an official tournament consisting of 4 rounds of 18 holes. EI and AEE were determined using HR wrist monitoring (Whoop Strap 2.0). We calculated the percentage of HRmax (%HRmax) and the percentage of HRres (%HRres) and the AEE in kcal/min using Keytel's formula. The calculated mean %HRmax and %HRres for the study population were 56.4% ± 1.8% and 40.5% ± 2.6%, respectively. Considering American College of Sports Medicine guidelines, these average percentages correspond to a moderate EI. The average caloric expenditure was 5.4 ± 0.4 kcal/min and 1555.8 ± 157.8 kcal per round considering an average golf round duration of 288.3 ± 19.5 minutes. A professional player's golf round is moderate physical activity. The AEE of this activity was equal to 5.4 cal/min, which is moderate energy consumption. These data could help golf coaches and conditioning coaches to have a better understanding of the load placed on golfers during tournaments.

Validation of Metabolic Models for Estimation of Energy Expenditure During Isolated Concentric and Eccentric Muscle Contractions.

Musculoskeletal modeling uses metabolic models to estimate energy expenditure of human locomotion. However, accurate estimation of energy expenditure is challenging, which may be due to uncertainty about the true energy cost of eccentric and concentric muscle contractions. The purpose of this study was to validate three commonly used metabolic models, using isolated isokinetic concentric and eccentric knee extensions/flexions. Five resistance-trained adult males (25.6 ± 2.4 year, 90.6 ± 7.5 kg, 1.81 ± 0.09 m) performed 150 repetitions at four different torques in a dynamometer. Indirect calorimetry was used to measure energy expenditure during these muscle contractions. All three models underestimated the energy expenditure (compared with indirect calorimetry) for up to 55.8% and 78.5% for concentric and eccentric contractions, respectively. Further, the coefficient of determination was in general low for eccentric contractions (R2 < 0.46) indicating increases in the absolute error with increases in load. These results show that the metabolic models perform better when predicting energy expenditure of concentric contractions compared with eccentric contractions. Thus, more knowledge about the relationship between energy expenditure and eccentric work is needed to optimize the metabolic models for musculoskeletal modeling of human locomotion.

TG/HDL Ratio Is an Independent Predictor for Estimating Resting Energy Expenditure in Adults with Normal Weight, Overweight, and Obesity

Factors that determine resting energy expenditure (REE) remain under investigation, particularly in persons with a high body mass index (BMI). The accurate estimation of energy expenditure is essential for conducting comprehensive nutrition assessments, planning menus and meals, prescribing weight and chronic disease interventions, and the prevention of malnutrition. This study aimed to: (a) determine the contribution of cardiometabolic biomarkers to the inter-individual variation in REE in persons categorized by BMI; and (b) assess the contribution of these biomarkers in the prediction of REE when persons of varying BMI status were categorized by their glycemic and metabolic syndrome status. Baseline data from 645 adults enrolled in diet intervention trials included REE measured by indirect calorimetry, body composition by dual energy X-ray absorptiometry, anthropometrics, and cardiometabolic biomarkers. Multivariate linear regression modeling was conducted to determine the most parsimonious model that significantly predicted REE by BMI category, metabolic syndrome status, and glycemic status. Modeling with the traditional predictors (age, sex, height, weight) accounted for 58-63% of the inter-individual variance in REE. When including age, sex, height, weight and fat-free mass as covariates, adding TG/HDL to regression modeling accounted for 71-87% of the variance in REE. The finding that TG/HDL is an independent predictor in estimating REE was further confirmed when participants were categorized by metabolic syndrome status and by glycemic status. The clinical utility of calculating the TG/HDL ratio not only aids health care providers in identifying patients with impaired lipid metabolism but can optimize the estimation of REE to better meet therapeutic goals for weight and disease management.

Assessment of energy expenditure in postpartum women with prior gestational diabetes mellitus: Subjective versus objective methods

Background: Regular physical activity (PA) attenuates the development of type 2 diabetes mellitus in postpartum women with prior gestational diabetes mellitus (GDM). PA is assessed by calculating the energy expenditure (EE), thus accurate estimation of EE is crucial. International PA Questionnaire (IPAQ) and activity diary are widely used subjective methods while pedometer provides objective evidence of PA and EE. Aims and Objectives: This study aims to compare subjective and objective methods of assessing EE as a measure of PA in postpartum women with a history of GDM. Materials and Methods: Fifty postpartum women who had a history of GDM were advised to maintain activity diaries and pedometer readings for 1 week each month up to 1 year and IPAQ was administered at the end of 6 and 12 months. EE was calculated by all three methods separately and Bland-Altman plots were used to assess the level of agreement between IPAQ and the other tools at 6 and 12 months. Friedman test was performed to compare estimated EE between the three methods. Results: There was a poor agreement between the subjective and objective methods. According to Bland-Altman plot results, the highest mean difference was observed between IPAQ and pedometer at 12 months (mean difference was 75.9). This observation indicated an underestimation of pedometer values at 12 months. EE assessed by two subjective methods were statistically comparable at both time points whereas, EE assessed by pedometer was significantly lower (P&lt;0.017) compared to self-reported tools. Conclusion: Assessment of EE by subjective methods which are based on personal judgment may cause overestimation of PA. In contrast, pedometer readings are likely to underestimate the PA as it only reports kinetics activities. Therefore, a combination of subjective and objective methods is recommended to ameliorate the reliability and validity.

Depth-Camera Based Energy Expenditure Estimation System for Physical Activity Using Posture Classification Algorithm.

Insufficient physical activity is common in modern society. By estimating the energy expenditure (EE) of different physical activities, people can develop suitable exercise plans to improve their lifestyle quality. However, several limitations still exist in the related works. Therefore, the aim of this study is to propose an accurate EE estimation model based on depth camera data with physical activity classification to solve the limitations in the previous research. To decide the best location and amount of cameras of the EE estimation, three depth cameras were set at three locations, namely the side, rear side, and rear views, to obtain the kinematic data and EE estimation. Support vector machine was used for physical activity classification. Three EE estimation models, namely linear regression, multilayer perceptron (MLP), and convolutional neural network (CNN) models, were compared and determined the model with optimal performance in different experimental settings. The results have shown that if only one depth camera is available, optimal EE estimation can be obtained using the side view and MLP model. The mean absolute error (MAE), mean square error (MSE), and root MSE (RMSE) of the classification results under the aforementioned settings were 0.55, 0.66, and 0.81, respectively. If higher accuracy is required, two depth cameras can be set at the side and rear views, the CNN model can be used for light-to-moderate activities, and the MLP model can be used for vigorous activities. The RMSEs for estimating the EEs of standing, walking, and running were 0.19, 0.57, and 0.96, respectively. By applying the different models on different amounts of cameras, the optimal performance can be obtained, and this is also the first study to discuss the issue.

The accuracy of fitness watches for the measurement of heart rate and energy expenditure during moderate intensity exercise.

As new technology emerges and updated fitness watches are released to the market, it is important to examine their accuracy. The aim of the current study was to examine the accuracy of three commercially available activity trackers in assessing heart rate (HR) and energy expenditure (EE) during moderate intensity exercise. Twenty healthy participants (Age: 20.5±0.7 yrs., Ht: 173.4±10.8 cm, BM: 72.8±13.9 kg, BMI: 24.0±2.5 kg/m2) wore two fitness watches (FB: Fitbit VersaTM, San Francisco, CA, USA; and PI: Polar IgniteTM, Polar Electro Oy, Kempele, Finland) and a chest-worn HR monitor (PTP: Polar TeamPro SensorTM, Polar Electro) during a 12-minute exercise protocol at incremental speeds. An electrocardiogram (ECG) and indirect calorimetry were used as criterion measures for HR and EE. Mean absolute percentage error (MAPE) was calculated to determine measurement error. The MAPE values for HR were 11.6±8.7% for the FB, 11.0±10.0% for the PI, and 6.3±5.2% for the PTP. For EE, MAPE values were 9.6±7.2% for the FB, 16.7±19.6% for the PI and 13.8±13.0% for the PTP. Fitness watches relying on optical measures of HR underestimate HR compared to criterion measures during moderate intensity exercise. Despite providing a more accurate measure of HR, a chest-worn monitor does not provide a more accurate estimate of EE compared to fitness watches. The Fitbit provided the most accurate measure of EE when compared to the Polar Ignite watch and chest-worn device.

Resting Energy Expenditure during Breastfeeding: Body Composition Analysis vs. Predictive Equations Based on Anthropometric Parameters.

Accurate estimation of energy expenditure in a breastfeeding woman is crucial for maintaining the proper nutritional status of the woman and healthy development of the infant. The current literature does not contain data regarding resting energy expenditure (REE) in breastfeeding women. Using mathematical equations is the most common method of REE assessment. However, due to changes in metabolism and body composition during pregnancy and lactation, the mathematical equations used among the general population may not apply. The aim of this study was to evaluate the resting energy expenditure of exclusively breastfeeding women by using body composition analysis–estimated REE (eREE) and to provide the most appropriate predictive equations–predicted REE (pREE) based on anthropometric parameters to estimate it. This was a pilot study with 40 exclusively breastfeeding women. Height and weight were measured and body composition analysis was performed. We predicted REE using fourteen self-selected equations, based on anthropometric parameters and/or age, and/or sex. The median eREE was 1515.0 ± 68.4 kcal (95% Cl, 1477–1582 kcal) and the pREE ranged from 1149.7 kcal (95% Cl, 1088.7–1215.0) by Bernstein et al., to 1576.8 kcal (95% Cl, 1479.9–1683.4), by Müller et al. Significant differences between eREE and all pREE were observed (p < 0.001, except Korth et al. equations). The Müller et al. equation was the most accurate with the smallest individual variation. All predictive equations showed low agreement, and in most cases, the results were underestimated. These findings indicate the need for further studies to propose more suitable methods to determine the energy requirements for breastfeeding women.

Physiological responses during a 25-km time trial in elite wheelchair racing athletes.

Observational study. To characterize the cardiorespiratory and metabolic response of elite wheelchair racing (WCR) athletes during a 25 km, field-based time trial. University laboratory and field racing course in Urbana, Illinois, USA. Seven elite WCR athletes (4 men/3 women) with spinal cord injury completed an incremental exercise test to exhaustion on a computerized wheelchair roller system to determine peak cardiorespiratory capacity in the laboratory. The athletes then completed a long-distance, field-based time trial (i.e., 25 km) within 5 days. Energy expenditure was measured continuously during the time trial with a portable metabolic unit. Blood samples were collected to determine blood lactate and glucose concentrations. Core temperature was measured using an ingestible sensor thermistor. Five participants completed the long-distance time trial with usable cardiorespiratory data. Median heart rate and oxygen consumption during the time trial was 93.6% and 76.6% of peak values, respectively. Median energy expenditure was 504.6 kcal/h. There was a significant increase in blood lactate concentration from 0.7 to 4.0 mmol/L after the time trial (p = 0.03). There were no changes in blood glucose concentrations after the time trial (p = 0.27). Lastly, core temperature significantly increased from 37.1 at baseline to 38.7 °C immediately after the time trial (p = 0.01). Elite WCR athletes sustained a high exercise intensity that was consistent across the long-distance time trial, and exercise intensity outcomes were generally lower than those documented for elite able-bodied long-distance athletes in other studies. Our findings provide accurate estimates of energy expenditure that can be used to design effective training and racing strategies for elite WCR athletes.

Validity of predictive equations for resting metabolic rate in healthy older adults

Accurate estimation of energy expenditure in older people is important for nutritional support. The current literature contains controversial or inconsistent data regarding the resting metabolic rate (RMR, or basal metabolic rate) in older adults, including the relationship between the RMR and ethnicity. Little information about the RMR in healthy Asian older adults is available. This study was performed to examine the RMR in healthy Japanese older adults and compare it with previously established 16 equations. Thirty-two community-dwelling, healthy, and active elderly Japanese adults were enrolled (age, 64-87 years; 14 men, 18 women; mean height, 154.9±8.9cm; mean weight, 53.5±9.1kg; mean body mass index, 22.2±2.5kg/m2). The RMR was measured by indirect calorimetry. The measured RMR was compared among 16 equations. Correlation analysis, a paired t test, and a Bland-Altman plot were used to assess the agreement among the equations. The average RMR was 1132±178kcal/day with 2233±437kcal/day average total energy expenditure (TEE) measured by doubly labeled water (DLW). The smallest bias was established by De Lorenzo etal.'s equation as bias ±1.96SD=4±121kcal/day. De Lorenzo etal. and Ikeda etal.'s equations had no significant average bias both in men and women (P>0.05). The 1.96SD of bias in six equations was within 160kcal/day. In contrast, residuals between the measured and predicted RMR were largely correlated with the RMR in four equations. A sex-related difference in the mean bias was observed in many equations. Although the average Japanese healthy older adult has a shorter stature and lower weight than older adults in the Western population, the current data suggest that a similar predictive equation for the RMR can be applied to both Japanese and Western older adults. This study demonstrate that the De Lorenzo etal.'s or Ikeda's equation may be useful for estimating RMR in the community-dwelling, healthy, and active elderly Japanese adults without any systematic bias.

Estimation of Energy Expenditure during Treadmill Exercise via Thermal Imaging

Noninvasive imaging of oxygen uptake may provide a useful tool for the quantification of energy expenditure during human locomotion. A novel thermal imaging method (optical flow) was validated against indirect calorimetry for the estimation of energy expenditure during human walking and running. Fourteen endurance-trained subjects completed a discontinuous incremental exercise test on a treadmill. Subjects performed 4-min intervals at 3, 5, and 7 km·h (walking) and at 8, 10, 12, 14, 16, and 18 km·h (running) with 30 s of rest between intervals. Heart rate, gas exchange, and mean accelerations of ankle, thigh, wrist, and hip were measured throughout the exercise test. A thermal camera (30 frames per second) was used to quantify optical flow, calculated as the movements of the limbs relative to the trunk (internal mechanical work) and vertical movement of the trunk (external vertical mechanical work). Heart rate, gross oxygen uptake (mL·kg·min) together with gross and net energy expenditure (J·kg·min) rose with increasing treadmill velocities, as did optical flow measurements and mean accelerations (g) of ankle, thigh, wrist, and hip. Oxygen uptake was linearly correlated with optical flow across all exercise intensities (R = 0.96, P < 0.0001; V˙O2 [mL·kg·min] = 7.35 + 9.85 × optical flow [arbitrary units]). Only 3-4 s of camera recording was required to estimate an optical flow value at each velocity. Optical flow measurements provide an accurate estimation of energy expenditure during horizontal walking and running. The technique offers a novel experimental method of estimating energy expenditure during human locomotion, without use of interfering equipment attached to the subject.

Metabolic Power Method: Underestimation of Energy Expenditure in Field-Sport Movements Using a Global Positioning System Tracking System.

The purpose of this study was to assess the validity of a global positioning system (GPS) tracking system to estimate energy expenditure (EE) during exercise and field-sport locomotor movements. Twenty-seven participants each completed a 90-min exercise session on an outdoor synthetic futsal pitch. During the exercise session, they wore a 5-Hz GPS unit interpolated to 15 Hz and a portable gas analyzer that acted as the criterion measure of EE. The exercise session was composed of alternating 5-minute exercise bouts of randomized walking, jogging, running, or a field-sport circuit (×3) followed by 10 min of recovery. One-way analysis of variance showed significant (P < .01) and very large underestimations between GPS metabolic power- derived EE and oxygen-consumption (VO2) -derived EE for all field-sport circuits (% difference ≈ -44%). No differences in EE were observed for the jog (7.8%) and run (4.8%), whereas very large overestimations were found for the walk (43.0%). The GPS metabolic power EE over the entire 90-min session was significantly lower (P < .01) than the VO2 EE, resulting in a moderate underestimation overall (-19%). The results of this study suggest that a GPS tracking system using the metabolic power model of EE does not accurately estimate EE in field-sport movements or over an exercise session consisting of mixed locomotor activities interspersed with recovery periods; however, is it able to provide a reasonably accurate estimation of EE during continuous jogging and running.

Feasibility Of A Novel Accelerometer Attachment-site For The Measurement Of Energy Expenditure

Energy expenditure (EE) is an important behavioral contributor to the current obesity pandemic, therefore accurate measures of EE are needed. It has been suggested accelerometers are affordable devices to provide valid estimates of EE, however, the controversy over which attachment-site for an accelerometer is the most accurate and feasible/compliant for this purpose continues. The Waist and Wrist sites are commonly used in research and both have well-documented strengths and weaknesses. Recently, there has been a growing interest in the Chest attachment-site as it has the advantages of being hidden from view but close to the center of mass and therefore the potential to provide accurate estimates of EE. PURPOSE: To determine if wearing the GT3X+ at the Chest (like a neck-pendant, beneath clothing) can provide valid estimates of EE. METHODS: A group of 10 young adults (20-26 years) wore GT3X+ accelerometers simultaneously at their Waist, Wrist, and Chest, whilst undertaking bouts of slow, medium and fast speeds (4.0, 6.4 and 8.0 kph, respectively) of walking and jogging on a treadmill. EE was measured by a primary criterion (indirect calorimetry, Cortex Metamax 3B), and a secondary criterion (BodyMedia SenseWear). For each attachment-site, a Friedman test was used to compare GT3X+ estimated EE (3-min of steady-state at each speed) against the primary criterion EE. Pairwise comparisons were performed with a Bonferroni correction for multiple comparisons. RESULTS: The median total EE measured by GT3X+ was significantly different across the attachment-sites (p < 0.005). Post hoc analysis revealed significant differences between the primary criterion EE (17.2kcal) and GT3X+ measured EE at the Wrist (9.0kcal, p < 0.005) but not the Chest (16.5kcal, p=0.79) or the Waist (16.2kcal, p=0.25). The Chest site (16.5kcal) and secondary criterion (15.5kcal, p= 0.66) were also similar. CONCLUSIONS: Preliminary results of this study suggest that the novel Chest attachment-site is able to provide valid estimates of EE compared to both the primary and secondary criterion. We are currently undertaking further data collection to increase our sample size. Funding: Hong Kong University Seed Fund for Basic Research.

Validation of a wireless accelerometer network for energy expenditure measurement

ABSTRACTThe purpose of this study was to validate a wireless network of accelerometers and compare it to a hip-mounted accelerometer for predicting energy expenditure in a semi-structured environment. Adults (n = 25) aged 18–30 engaged in 14 sedentary, ambulatory, exercise, and lifestyle activities over a 60-min protocol while wearing a portable metabolic analyser, hip-mounted accelerometer, and wireless network of three accelerometers worn on the right wrist, thigh, and ankle. Participants chose the order and duration of activities. Artificial neural networks were created separately for the wireless network and hip accelerometer for energy expenditure prediction. The wireless network had higher correlations (r = 0.79 vs. r = 0.72, P < 0.01) but similar root mean square error (2.16 vs. 2.09 METs, P > 0.05) to the hip accelerometer. Measured (from metabolic analyser) and predicted energy expenditure from the hip accelerometer were significantly different for the 3 of the 14 activities (lying down, sweeping, and cycle fast); conversely, measured and predicted energy expenditure from the wireless network were not significantly different for any activity. In conclusion, the wireless network yielded a small improvement over the hip accelerometer, providing evidence that the wireless network can produce accurate estimates of energy expenditure in adults participating in a range of activities.

Constructing energy expenditure regression model using heart rate with reduced training time.

Accurate estimation of energy expenditure (EE) is a key enabler for many applications of healthcare and wellness. Heart rate (HR) based EE estimation methods typically require extensive training time to establish a relationship between HR and EE. In this work, we propose a method where just the few most representative EE-HR data pairs are used to train the estimation model. Furthermore, we present a systematical methodology based on the ranking of the correlation coefficients between EE and HR to find the least amount of EE-HR data pairs required for training while satisfying the constraint of estimation accuracy. During the experimental evaluation, while the study participants walk and run on a treadmill, our method is compared to three different training paradigms: training the EE-HR model 1) using all available data collected during the experiment, 2) using the EE-HR data only during speed changes (or during monotonic HR changes) and 3) using the EE-HR data pairs collected during constant speed. The results show that our method could maintain a comparable EE estimation performance as shown by only 2~4% changes on the coefficient of variation of root-mean-squared error (CV(RMSE)) for the testing dataset while saving nearly 91-97% training time for each individual.

Data correction for seven activity trackers based on regression models.

Using an activity tracker for measuring activity-related parameters, e.g. steps and energy expenditure (EE), can be very helpful in assisting a person's fitness improvement. Unlike the measuring of number of steps, an accurate EE estimation requires additional personal information as well as accurate velocity of movement, which is hard to achieve due to inaccuracy of sensors. In this paper, we have evaluated regression-based models to improve the precision for both steps and EE estimation. For this purpose, data of seven activity trackers and two reference devices was collected from 20 young adult volunteers wearing all devices at once in three different tests, namely 60-minute office work, 6-hour overall activity and 60-minute walking. Reference data is used to create regression models for each device and relative percentage errors of adjusted values are then statistically compared to that of original values. The effectiveness of regression models are determined based on the result of a statistical test. During a walking period, EE measurement was improved in all devices. The step measurement was also improved in five of them. The results show that improvement of EE estimation is possible only with low-cost implementation of fitting model over the collected data e.g. in the app or in corresponding service back-end.

Personalized cardiorespiratory fitness and energy expenditure estimation using hierarchical Bayesian models

Accurate estimation of energy expenditure (EE) and cardiorespiratory fitness (CRF) is a key element in determining the causal relation between aspects of human behavior related to physical activity and health. In this paper we estimate CRF without requiring laboratory protocols and personalize energy expenditure (EE) estimation models that rely on heart rate data, using CRF. CRF influences the relation between heart rate and EE. Thus, EE estimation based on heart rate typically requires individual calibration. Our modeling technique relies on a hierarchical approach using Bayesian modeling for both CRF and EE estimation models. By including CRF level in a hierarchical Bayesian model, we avoid the need for individual calibration or explicit heart rate normalization since CRF accounts for the different relation between heart rate and EE in different individuals. Our method first estimates CRF level from heart rate during low intensity activities of daily living, showing that CRF can be determined without specific protocols. Reference VO2max and EE were collected on a sample of 32 participants with varying CRF level. CRF estimation error could be reduced up to 27.0% compared to other models. Secondly, we show that including CRF as a group level predictor in a hierarchical model for EE estimation accounts for the relation between CRF, heart rate and EE. Thus, reducing EE estimation error by 18.2% on average. Our results provide evidence that hierarchical modeling is a promising technique for generalized CRF estimation from activities of daily living and personalized EE estimation.

Using Smartphone Sensors for Improving Energy Expenditure Estimation.

Energy expenditure (EE) estimation is an important factor in tracking personal activity and preventing chronic diseases, such as obesity and diabetes. Accurate and real-time EE estimation utilizing small wearable sensors is a difficult task, primarily because the most existing schemes work offline or use heuristics. In this paper, we focus on accurate EE estimation for tracking ambulatory activities (walking, standing, climbing upstairs, or downstairs) of a typical smartphone user. We used built-in smartphone sensors (accelerometer and barometer sensor), sampled at low frequency, to accurately estimate EE. Using a barometer sensor, in addition to an accelerometer sensor, greatly increases the accuracy of EE estimation. Using bagged regression trees, a machine learning technique, we developed a generic regression model for EE estimation that yields upto 96% correlation with actual EE. We compare our results against the state-of-the-art calorimetry equations and consumer electronics devices (Fitbit and Nike+ FuelBand). The newly developed EE estimation algorithm demonstrated superior accuracy compared with currently available methods. The results were calibrated against COSMED K4b2 calorimeter readings.

Validation of the SenseWear Pro3 Armband using an incremental exercise test.

The purpose of this study is to assess the accuracy of a multisensor activity monitor in measuring energy expenditure (EE) using an incremental exercise test. The SenseWear Pro3 Armband (SWA) has been shown a valid and practical tool in measuring the EE of daily living activities but shows significant errors in estimating EE of activities of vigorous and very vigorous intensity. Recently, a new algorithm was developed for SWA (SWA v5.2). The aim of this study was to validate the new SWA algorithm in assessing EE using an incremental exercise test. Data were obtained from 44 students (age, 21.1 ± 1.4 years). During an incremental running test, EE estimates of the SWA were compared with the indirect calorimetry (IC). A paired t-test, Bland and Altman plots, and correlation coefficients were used as statistical techniques. Compared with the IC, the SWA showed no significant difference when walking at 1.5 m·s (p = 0.901) but significantly underestimated EE at higher speeds starting from a jogging speed of 1.67 m·s for men and from a running speed of 2 m·s for women. The underestimation increased significantly with increasing intensity in both male and female participants. The accurate measurement of EE at moderate intensity makes the monitor a valid tool in estimating daily life activities. However, our study results question the applicability of the monitor for individuals willing to monitor their EE during vigorous or very vigorous exercise intensities because a significant underestimation of the new algorithm is still present. In future, the EE algorithm needs to be further adjusted to give an accurate estimate of EE for high-intensity exercises.