See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/49658583
Accurate Prediction of Energy Expenditure Using
a Shoe-Based Activity Monitor
Article in Medicine and science in sports and exercise · December 2010
DOI: 10.1249/MSS.0b013e318206f69d · Source: PubMed
CITATIONS
READS
24
94
3 authors:
Nadezhda Sazonova
Raymond C Browning
21 PUBLICATIONS 255 CITATIONS
78 PUBLICATIONS 1,206 CITATIONS
Independent Researcher
SEE PROFILE
Nike Inc.
SEE PROFILE
Edward Sazonov
University of Alabama
134 PUBLICATIONS 1,357 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
At Nike, managing a team of scientists who conduct sport science research. View project
Shoe-based wearable sensors View project
All content following this page was uploaded by Edward Sazonov on 18 January 2017.
The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
D
. . . Published ahead of Print
TE
Accurate Prediction of Energy Expenditure Using
a Shoe-Based Activity Monitor
Nadezhda Sazonova1, Raymond C. Browning2, and Edward Sazonov1
1
C
EP
Department of Electrical and Computer Engineering, University of Alabama, Tuscaloosa,
AL2Department of Health and Exercise Science, Colorado State University, Fort Collins, CO
A
C
Accepted for Publication: 9 November 2010
Medicine & Science in Sports & Exercise® Published ahead of Print contains articles in unedited
manuscript form that have been peer reviewed and accepted for publication. This manuscript will undergo
copyediting, page composition, and review of the resulting proof before it is published in its final form.
Please note that during the production process errors may be discovered that could affect the content.
Copyright © 2010 American College of Sports Medicine
Medicine & Science in Sports & Exercise, Publish Ahead of Print
DOI: 10.1249/MSS.0b013e318206f69d
Accurate Prediction of Energy Expenditure Using a Shoe-Based Activity Monitor
Nadezhda Sazonova1, Raymond C. Browning2 and Edward Sazonov1
Department of Electrical and Computer Engineering, the University of Alabama, USA
2
Department of Health and Exercise Science, Colorado State University, USA
Nadezhda Sazonova
TE
Address for Correspondence:
D
1
Department of Electrical and Computer Engineering
317 Houser Hall
C
EP
the University of Alabama
Tuscaloosa, AL 35487-0286
phone: (205) 348-6351; fax: (205) 348-6959; Email:
[email protected]
Prediction of Energy Expenditure Using Shoe
C
Running title:
A
This work was supported in part by the University of Colorado Technology Transfer Office Proof
of Concept Grant, and NIH grant 1R43DK083229
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
ABSTRACT
Purpose: The aim of this study was to develop and validate a method for predicting energy
expenditure (EE) using a footwear-based system with integrated accelerometer and pressure
sensors. Methods: We developed a footwear-based device with an embedded accelerometer and
insole pressure sensors for the prediction of energy expenditure. The data from the device can be
D
used to perform accurate recognition of major postures and activities and to estimate EE using
the acceleration, pressure and posture/activity classification information in a branched algorithm
TE
without the need for individual calibration. We measured EE via indirect calorimetry as sixteen
adults (BMI: 19-39 kg∙m-2) performed various low-to-moderate intensity activities and compared
measured vs. predicted EE using several models based on the acceleration and pressure signals.
C
EP
Results: Inclusion of pressure data resulted in better accuracy of EE prediction during static
postures such as sitting and standing. The activity-based branched model that included predictors
from accelerometer and pressure sensors (BACC-PS) achieved the lowest error (e.g. root mean
squared error (RMSE) of 0.69 METs) compared to the accelerometer-only based branched model
BACC (RMSE of 0.77 METs) and non-branched model (RMSE of 0.94-0.99 METs).
Comparison of EE prediction models using data from both legs vs. models using data from a
C
single leg indicate that only one shoe need to be equipped with sensors. Conclusion: These
A
results suggest that foot acceleration combined with insole pressure measurement, when used in
an activity-specific branched model, can accurately estimate the energy expenditure associated
with common daily postures and activities. The accuracy and unobtrusiveness of a footwearbased device may make it an effective physical activity monitoring tool.
Keywords: indirect calorimetry; pressure sensors; accelerometry; wearable sensors; shoe
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
INTRODUCTION
Paragraph Number 1 Physical activity (PA) levels and the energy expenditure (EE) associated
with physical activity influence human health (33). As a result, individuals are advised to
participate in programs that promote increased energy expenditure (EE) via exercise, physical
activity and changes in posture allocation (e.g. less sitting) (16). Accurately quantifying levels of
D
physical activity (PA) and associated EE in adults and children will provide insights into the
dose-response relationship between PA/EE and health outcomes, allow evaluation of the
TE
effectiveness of interventions that aim to increase PA/EE and aid in treating metabolic disorders
associated with obesity. Monitoring physical activity patterns objectively (e.g. via accelerometry)
can improve PA/EE estimates, but devices that can accurately estimate total daily and activity-
C
EP
specific EE are essential. For example, the magnitude of positive energy balance that results in
gradual weight gain is on the order of 25-100 kcal/day.
In addition, instruments that are
unobtrusive and easy to use may improve compliance and reduce limitations to physical activity
due to the device interfering with movement.
Paragraph Number 2 Accelerometry (ACC) has emerged as one of the most popular approaches
to EE prediction (6,10,12,15,25,28). Although useful, single accelerometers have one major
C
drawback in that they tend to significantly underestimate the energy cost of static postures such
A
as standing activities (e.g. household tasks) and non weight-bearing activities (e.g. cycling) (18).
As a result, they fail to explain a considerable portion of energy expenditure variability in daily
living tasks. One strategy to improve EE estimation has been to use multiple sensors, either
additional accelerometers or other types of sensors (e.g. heart rate) (7,16,29,30). For example,
combining heart rate and ACC has been shown to substantially improve the accuracy of energy
expenditure prediction (7,29,30), as has the use of multiple accelerometers (35). Recently,
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
several studies have demonstrated improved EE estimation with a single accelerometer by using
more sophisticated modeling approaches including Artificial Neural Networks (28), distributed
lag and spline modeling (11) and branched algorithms (12,13). Another way to achieve an
improvement in EE accuracy has been to use the ACC data to classify activity, which is used in
predictive models based on the type or intensity of the activity (6,12,28).
D
Paragraph Number 3 Heart rate monitoring and multiple ACCs are the most common ways
explored to supplement single ACC’s in energy expenditure prediction. Exploration of other
TE
approaches may lead to an improved prediction accuracy and greater convenience to a weight
management participant. Recently, we developed a wearable shoe-based device (26) which has
an embedded accelerometer and pressure sensors positioned in the insole. The main appeal of
C
EP
using the device for energy expenditure prediction is its potential accuracy, non-intrusiveness,
light weight and ease of use. We have developed a posture and activity recognition model for
this device which is able to achieve 98% accuracy in subject-independent classification of 6
major postures and activities (sitting, standing, walking, ascending stairs, descending stairs and
cycling) (27).
This enables an activity-specific branched approach to energy expenditure
prediction that may result in relatively good EE estimates for a variety of daily living tasks. The
C
inclusion of insole pressure sensors in the device also allows the exploration of whether the
A
intensity of physical activity may be correlated with range and frequency of foot pressure
changes and whether pressure data can supplement the accelerometer data for further
improvement in accuracy of predicting EE. Thus, we developed and validated a method for using
accelerometer and pressure sensors signals to predict EE conditioned on a specific activity group
and without need of individual calibration. Several studies reported using shoe-based sensors
(3,17,20), however, their research concentrated on detecting gait characteristics, rather than
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
posture/activity classification and energy expenditure estimation. There are also several
commercially distributes shoe-based systems (such as Pedar (31) and F-Scan (14)) which
incorporate pressure sensors in the insoles for the dynamic pressure measurements. Although
these systems have wide applicability such as kinetic analysis of gait, shoe research and design,
orthotic design, podiatry and sports biomechanics, they are not designed specifically for
D
posture/activity recognition and energy expenditure prediction. A study reported in (34) used an
array of 32 plantar pressure sensors to classify locomotion (walking, running and up/down
TE
stairs). Another study (32) estimated daily energy expenditure using a foot-contact pedometer but
did not attempt to classify postures or specific activities with the device. We introduce a shoebased device that will be the first in the area of footwear-based systems to be used for accurate
C
EP
posture/activity recognition and energy expenditure estimation.
Paragraph Number 4 The main purpose of this study was to test the overall feasibility of energy
expenditure prediction using a novel shoe-based device, in particular, we aimed to perform the
following tasks : 1) to compare the accuracy of EE prediction using this device vs. existing
methods using single ACC or ACC/HR sensors; 2) to compare the accuracy of prediction
performance of a model using accelerometer and pressure sensors signals vs. a model that uses
C
only accelerometer signal; 3) to validate the branched modeling approach for prediction of
A
energy expenditure for each specific posture and activity; 4) to evaluate the need of sensors to be
embedded in both shoes. We hypothesized that the combination of ACC and pressure data would
provide more accurate EE estimates compared to single ACC /EE methods and that sensors
would only be required in a single shoe.
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
METHODS
Subjects
Paragraph Number 5 Sixteen adult subjects participated in the study. The University
Institutional Review Board approved the study and each subject provided informed consent. In
order to test the device on a diverse population, we recruited participants who were lean to obese.
D
Based on self-report, participants weight was stable (<2 kg weight fluctuation) over the previous
6 months. Individuals who were healthy, non-smokers, and sedentary to moderately active (< 2-3
participate in the study.
TE
bouts of exercise/wk or participation in any sporting activities < 3 hr/wk) were invited to
Pregnant women and those who had impairments that prevented
physical activity were excluded. The physical characteristics of participants are shown in Table
Study design
C
EP
1.
Paragraph Number 6 Participants reported to the laboratory in a fasted state (>4 hours) for a
single three hour visit. Each participant was asked to perform a variety of postures/activities
while wearing a portable metabolic cart system and the appropriately sized shoe device with
embedded sensors.
The postures included sitting and standing and the activities included
C
walking, jogging, stair ascent/descent and cycling (Table 2). Each posture/activity trial was six
A
minutes in duration and subjects were allowed five minutes rest between trials. Trial order was
not randomized. Metabolic data was not collected during stair ascent/descent, as this activity
was performed in two-story stairwell which did not allow establishment of metabolic steadystate. As a result, we estimated EE as each participant performed 13 different activities from
four posture/activity groups (Sit, Stand, Walk/Jog and Cycle).
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
Paragraph Number 7 Participants were not restricted in the way they assumed postures
and or performed activities. Standing did not require any specialized equipment; a chair with a
rigid back was used for sitting; walking/jogging was performed on a motorized treadmill (Gait
Trainer 1, Biodex, Shirley, NY); cycling utilized a bicycle ergometer (Erogomedic 828E,
movements (e.g. crossing legs or shifting weight).
EE measurement
D
Monark, Sweden). During the fidgeting trials, subjects were allowed to make small, normal leg
TE
Paragraph Number 8 To determine metabolic rate and associated EE during each trial, we
measured the rates of oxygen consumption (VO2) and carbon dioxide production (VCO2) using a
portable open circuit respirometry system (Oxycon Mobile, Viasys, Yorba Linda, CA). Before
C
EP
the experimental trials, we calibrated the system with known gas concentrations and volumes.
For each trial, we allowed four minutes for subjects to reach steady state (no significant increase
in VO2 during the final two minutes and a respiratory exchange ratio (RER) <1.0) and calculated
the average VO2 and VCO2 (ml/sec) during minutes 4-6 of each trial. We calculated gross
metabolic rate (W/kg) from VO2 and VCO2 using a standard equation (6). Energy expenditure
was then calculated from VO2 and RER.
C
Movement and foot pressure measurement.
A
Paragraph Number 9 The sensor data for this study were collected by a wearable sensor system
embedded into shoes (Fig. 1). Each shoe incorporated five force-sensitive resistors embedded in
a flexible insole and positioned under the critical points of contact: heel, metatarsal bones and the
great toe (hallux). The acceleration data were collected from a 3-dimensional MEMS
accelerometer positioned on the back of the shoe. The goal of the accelerometer was to detect
orientation of the shoe with respect to gravity, to characterize the motion trajectory and to help
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
characterize the amount of movement in a specific posture or activity. Pressure and acceleration
data were sampled at 25Hz and sent over a wireless link to the base computer.
Paragraph Number 10 The wireless system used for data acquisition was based on Wireless
Intelligent Sensor and Actuator Network (WISAN) (21). The battery, power switch and the
WISAN board were installed at the back of the shoe as shown on Fig. 1(b). The sensor system
D
was lightweight (<40g) and created no visible interference with the motion patterns in subjects.
Model
TE
Paragraph Number 11 For the model construction we used a group rather than individual
approach: the data used for training were pooled from several subjects and such model was then
tested on the validation set which included data from subject(s) that were not in the training set.
C
EP
For each posture and activity the sensor data were collected during a 1 minute interval in which
subjects were in metabolic steady state (minutes 4-6 of each trial). Each one minute recording
resulted in approximately 1500 (25Hz∙60s) points of pressure and acceleration data per channel.
For the 16 subjects who participated in the study there were a total of 208 such recordings.
The following data were available for each recording:
response variable: energy expenditure, EE, kcal·min-1;
•
anthropometric measurements (weight, height, BMI, age, gender, shoe size);
•
triaxial
C
•
signals:
superior-inferior
acceleration
(Acc1),
medial-lateral
A
accelerometer
acceleration (Acc2), anterior-posterior acceleration(Acc3);
•
pressure sensors signals: heel (Sens1), 3rd meta (Sens2), 1st meta (Sens3), 5th meta (Sens
4), and hallux (Sens5);
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
Paragraph Number 12 To validate the branching approach, energy expenditure prediction was
performed as a two-step process, with the step one being classification of postures/activities into
one of the four groups: “Sit”, “Stand”, “Walk” and “Cycle”; and step two being prediction of
energy expenditure using one of the four regression models built for a given posture/activity
group. Each 1-minute interval of sensor data was first classified as belonging to one of the four
D
activity groups using our earlier developed algorithm for posture/activity recognition (27). The
same sensor data from each 1-minute interval were consequently used for training and validation
TE
of one of the four regression models for predicting energy expenditure. Thus, the branching
approach involved constructing four branch models: “Sit”, “Stand”, “Walk”, “Cycle” contingent
upon prior classification of every 1-min recording into one of these groups for training or
C
EP
validation. Another major goal was to justify the use of the pressure sensors (in addition to
accelerometer) in EE prediction. This led to the development of the following four models to
predict EE in kcal·min-1 using predictors described above:
1. BACC-PS. This model was branched by activity and consisted of four separate branch
models (“Sit”, “Stand”, “Walk”, “Cycle”). The predictors included anthropometric
measurements, accelerometer and pressure sensors as predictors;
C
2. BACC. This branched model also consisted of four separate branch models (“Sit”, “Stand”,
A
“Walk”, “Cycle”) and included anthropometric measurements and accelerometer data as
predictors but the pressure data were not used;
3. ACC-PS. This was a non-branched model (no activity classification) that used the same
predictors as BACC-PS.
4. ACC. This was a non-branched model using the same predictors as BACC.
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
The purpose of constructing different models was to investigate if the performance is improved
by branching the model (i.e. classifying the activity) and also by including predictors derived
from pressure signals.
Paragraph Number 13 Accelerometer and pressure sensors signals expressed in ADC units (the
signals were digitized by a 12-bit analog-to-digital converter) were preprocessed to extract
D
meaningful metrics to be used as predictors for the model. For both pressure and acceleration
sensors all of the following metrics were extracted and tested for the inclusion into each model as
TE
predictors: coefficient of variation (cv); standard deviation (std); number of “zero crossings” (zc),
i.e. number of times the signal crosses its median normalized by the signal's length; entropy H of
the distribution X of signal values (ent) computed as: H(X) = – Σ pk log pk , where pk is the
C
EP
relative frequency of values fallen into the kth interval (out of 20 equally sized intervals) in the
sample distribution of signal values. These metrics were selected for the following reasons.
Coefficient of variation and standard deviation of a signal should indicate the amount of motion
produced during recording, with the difference that coefficients of variations are affected by
signals mean value (for example, the gravitational component of acceleration) while standard
deviation is not. Number of median crossings is an indicator of the frequency of changes in the
C
signal, which is important to identify the intensity of motion (like speed of walking). Entropy
A
reflects the distribution of the signal across the range of its values and is a valuable predictor for
walking due to the fact that as speed of walking increases the time of feet ground contact
decreases relative to the swing time and, thus, signal values become more uniformly distributed
across the range, leading to an increased entropy.
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
Paragraph Number 14 For each model we used the derived metrics as possible predictors for the
ordinary least squares linear regression. The transformed predictors (log, inverse and square root)
and interactions (as products of 2 or more candidate predictors) were also considered as separate
linear terms within regression.
Paragraph Number 15 In branched models a separate model was constructed for each type
D
posture/activity: “Sit”, “Stand”, “Walk” and “Cycle”. For all branched (as collections of the four
separate branch models) and non-branched models selection of the most significant set of
TE
predictors was performed using the forward selection procedure. We used the “leave-one-out”
approach for cross-validation when training and predicting the EE for each type of activity for
every subject. For every left out subject all of the data related to this subject were removed from
C
EP
the training set. Model (coefficients) computed using the rest of the subjects sample was then
used to predict the EE for all trials of the left out subject. The best set of predictors had to
provide the best fit (by producing the maximum adjusted coefficient of determination, R 2adj and
the minimum Akaike Information Criterion, AIC) in the training step and the best predictive
performance (the minimum mean squared error, MSE and the minimum mean absolute error,
MAE ) in the validation step.
C
Paragraph Number 16 The input for the models was the data from sixteen subjects who had
A
complete metabolic and sensor data for all thirteen trials. In the “walk” activity group some
subjects did not have energy expenditure record (unable to achieve metabolic steady state while
jogging) or had no sensors signals recorded for some trials within this group, these 10 trials were
dropped from each model's input. An additional 1-minute recording for cycling activity for a
particular subject contained more than 50% of corrupted data due to sensor failure. This
recording was also dropped from the analysis. Thus, the sample size of the input data for each
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
model was 16·13 – 11 = 197 trials.
Paragraph Number 17 Measured and predicted energy expenditure values in kcal·min-1 for each
experiment were then converted to METs for both branched ACC-PS and ACC models and their
non-branched versions. The conversion from kcal·min -1 to METs was done by representing the
energy expenditure for any given epoch as a multiple of resting energy expenditure. We used
D
energy expenditure during quiet sitting as a valid estimate of resting metabolic rate for each
subject due to established convention (1,2). This conversion was performed to enable direct
TE
comparison of our results with those that have been recently published (9,12,28).
Paragraph Number 18 One of the goals of the analysis was to establish the need of using
sensors on both shoes. Several versions of the branched ACC-PS model (as a representative
C
EP
model) were constructed using accelerometer and pressure sensors data separately from each
shoe and both shoes together.
Statistics
Paragraph Number 19 The following performance assessment measures were computed for
each EE prediction model:
•
RMSEMET, the root mean squared error for energy expenditure prediction expressed in
C
METs. This error is computed as the difference between model predicted EE and the measured
EE for each trial.
A
•
95% confidence intervals for RMSEMET, computed as bootstrapping estimates by
generating 5000 samples of absolute errors (predicted minus actual energy expenditure) drawn
from the original sample, calculating RMSE MET for each such sample and computing bounds for
the middle 95% of the created population of RMSEMET’s.
•
ARD, the Average Relative Difference (signed):
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
ARD = mean((predEE - EE)/EE)
•
Bias, the mean difference between predicted and measured energy expenditure in METs:
bias = mean(predEE - EE)
•
Interval of agreement for prediction of energy expenditure in METs, calculated as given
in (1): (bias ± 2·SD(bias))
D
Paragraph Number 20 Bland-Altman plot analysis (4) was conducted to reveal any
systematic pattern of the error (calculated as the difference between predicted and measured EE)
TE
across the range of measurements and to assess the bias and interval of agreement for prediction
of EE.
Paragraph Number 21 Passing-Bablok regressions (a robust alternative to least squares
C
EP
regression) for all four models and for two units of prediction (kcal·min -1 and METs) were
constructed as described by Passing and Bablok (24). Passing-Bablok regression is best suited
for method comparison because it allows measurement error in both variables, does not require
normality of errors and is robust against outliers. In addition, Passing-Bablok regression
procedure estimates systematic errors in form of fixed (by testing if 95% CI includes 0) and
RESULTS
A
C
proportional bias (by testing if 95% CI includes 1).
Paragraph Number 22 Each raw signal of the accelerometer (3 sensors) and 5 pressure sensors
was represented by a vector of approximately 1500 measurements (25 measurements per
second). Sample raw signal for all 8 sensors is given in Fig., SDC1 for walking 2.5 mph activity.
Using the raw signal data, predictors for each model were computed by the following approach.
Metrics for the accelerometer and pressure sensors signals (cv, std, zc and ent) were computed
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
separately for the left and right shoe. Then the corresponding predictor values were formed as the
average of the left and right shoe metrics for the accelerometer signals, and as the maximum
value of the left or right shoe metrics for the same pressure sensor signals. The reason for the
maximum (rather than the average) in combining the left and right shoe metrics is that some of
the pressure sensors experienced occasional failure in three subjects. The signal from a failed
D
sensor would register as a constant zero value (no pressure), thus, using the maximum pressure
ensured that no data from failed sensor were used in training or validation. In particular, use of
TE
the maximum value resulted in the reduction of the corrupted data from 5% to around 1.7%.
Paragraph Number 23 To facilitate the branching approach, our automatic classification model
(26) for posture/activity recognition was applied to each of the 197 1-minute recordings to assign
C
EP
it into four activity groups (“Sit”, “Stand”, “Walk”, “Cycle”) for further construction and
validation of the corresponding branch models. For these data there was 100% rate of correct
classification among all 1-minute recordings with respect to the four activity groups.
Paragraph Number 24 Final linear regression coefficients for the branched ACC-PS and
branched ACC models after selection of the best set of predictors are reported in Table, SDC2
and Table, SDC3 respectively. The final non-branched ACC-PS and non-branched ACC model
C
regression coefficients are given in Table, SDC4. Among the anthropometric characteristics of
A
subjects used as possible predictors only Weight and BMI showed significance for energy
expenditure prediction for all models. In particular, gender-stratified models did not show any
improvement in the prediction performance. Similar effect was reported by previous studies
where gender has not been shown to improve EE estimates from accelerometry data (5,6,19).
The coefficients for all models were obtained by averaging the coefficients of the 11 runs (one
for each left out subject) of the OLS (Ordinary Least Squares) regression on the training sets.
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
Most of the coefficient of variations for coefficients of all of these models were within [0.07,
0.3], which suggests that the regression coefficients were highly stable.
Paragraph Number 25 Almost all coefficients for all models were highly stable over all runs as
given by low absolute values of coefficients of variation (CV). As can be expected, weight and
BMI always explain part of the variability of each model, while other physical characteristics
D
were highly correlated to weight variable and didn't add to the fit or the prediction performance
of either model. Results shown in Table 3 include performance comparison of the proposed
TE
BACC-PS model, BACC model, non-branched ACC-PS, non-branched ACC.
Paragraph Number 26 Bland-Altman plots (constructed for both EE, kcal·min -1 and EE, METs
prediction) for all four models are given in Fig. 2. Sub-figures (a)-(b) are Bland-Altman plots for
C
EP
branched models, sub-figures (c)-(d) are Bland-Altman plots for non-branched models. The
common characteristic for all these plots (models) is that the accuracy of prediction is slightly
better for small than for large EE values (i.e. better accuracy for sitting and standing).
Paragraph Number 27 Passing-Bablok regression analysis was conducted using Matlab
implementation (23) of the method described in (24). Examination of the presence of fixed
(intercept ≠ 0 if 95% CI does not contain 0) and proportional (slope ≠ 1 if 95% CI does not
C
contain 1) bias of the models showed that except for one case (non-branched ACC model which
A
showed fixed bias) none of the four models exhibited either kind of bias (see Table, SDC5,
examination of the presence of fixed and proportional bias and linearity). All ACC-PS models
(branched and non-branched) provided better prediction over the ACC models as indicated by
slope values closer to the unity than those of the ACC model (see Passing-Bablok regression
analysis in the supplemental materials). In addition, the branched model regression coefficients
appear to be more precise since they provided the narrower confidence intervals for both slope
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
and intercept than those for the non-branched models.
Paragraph Number 28 Linearity test indicated absence of linearity for all non-branched models
while for branched models linearity was always very strong (see Table, SDC5, examination of
the presence of fixed and proportional bias and linearity). Additional proof of the strength of
linear relationship between predicted and measured EE values is given by correlation and
D
concordance coefficients. There is clear tendency of both coefficients to increase from nonbranched to branched models and from ACC to ACC-PS models. Lack of linearity of the non-
TE
branched models is also noticeable in their Passing-Bablok regression plots (Fig. 3), which show
clear curvature in the scatter plots unlike in those of the branched models.
Paragraph Number 29 As a last step, we investigated the effect of inclusion of predictors from
We compared
C
EP
both shoes vs. a single shoe into the model using the BACC-PS model.
performance of the BACC-PS models that used the difference metrics derived from the
difference between signal form left and right shoe and/or the best selected set of predictors (as
metrics cv, std, zc and ent) computed separately for each shoe. Overall, models based on metrics
derived for both shoes perform slightly better (RMSE was within 0.68-0.70 METs) than single
shoe models (RMSE was 0.78 METs for left shoe-based model and 0.72 METs for right shoe-
C
based model), see Table, SDC6, comparison of BACC-PS model performance using predictors
A
from single shoe and both shoes. However, the RMSE values were still below those found for
BACC and the rest of the models. Also, the improvement of both-shoe over single-shoe models
can be attributed mostly to the lost of data due to sensors failure: both-shoe models were able to
mitigate the effect of the corrupted data by using the fact that simultaneous sensors failure on
both shoes was rare, and by applying either averaging or maximization to left and right shoe
sensor metrics. Thus, for all practical purposes single shoe models can be successfully used.
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
DISCUSSION
Paragraph Number 30 Our results suggest that a shoe-based device with embedded
accelerometer and pressure sensors can be used to accurately predict energy expenditure during
typical postures/physical activities. Such a device may be a useful tool for individuals interested
in weight management.
The combination of posture allocation/physical activity data (e.g.
modify or maintain energy balance.
D
minutes sitting and walking) with accurate estimates of EE can be used to help individuals
A shoe-based device may also be “invisible” and
TE
unobstrusive and lead to increased use, further facilitating weight management success.
Paragraph Number 31 The EE prediction accuracy of our device and branched model is similar
to recent studies that have used single accelerometers, multiple accelerometers and heart
C
EP
rate/accelerometer combinations. Choi et al. (11) used Actigraph accelerometers placed at the
hip, wrist and/or ankle and distributed lag and spline modeling to predict EE and reported RMSE
of ~0.6 kcal/min (0.5 METs) across a range of activities with the accelerometer mounted at the
ankle. Staudenmayer et al. used a single hip-mounted accelerometer (Actigraph) and an artificial
neural network to estimate EE of a variety of activities and reported an RMSE of 0.75 and 1.22
METs using activity and minute-by-minute estimates of EE, respectively (28). A study by
C
Crouter et al. that compared EE estimation using hip-mounted accelerometery vs. indirect
A
calorimetery reported systematic bias of 0.1 METs with 95% limits of agreement of (-1.4, 1.5)
METs (12). Although our results are not directly comparable to those from Staudenmayer et al.
and Crouter et al. as our subjects did not wear a hip mounted accelerometer and we did not
compare actual and predicted EE during the entire period of each trial, the similar RMSE values
suggest good agreement. Brage et al. used a device that measured heart rate and accelerometry
(Actiheart) to estimate EE and found that the RMSE was within [0.87, 1.11] METs during
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
walking/running activities (8). Thus, our results suggest the use of pressure and acceleration to
identify activity and predict EE are at least as accurate or better compared to other, recently
proposed methodologies.
Paragraph Number 32 Our results support the measurement of plantar pressure as a way to
improve EE prediction compared to a single accelerometer. As shown in Table 3, the inclusion
D
of pressure sensor metrics improved all prediction performance measures. In particular, RMSE
was reduced ~7% for branched models (0.77 to 0.69 METs) and non-branched models (0.99 to
TE
0.94 METs). There was also clear reduction in bias and the width of the interval of agreement
when comparing ACC to corresponding ACC-PS models. Because there are clear differences in
the magnitude and distribution of insole pressure across postures and activities, insole pressure
C
EP
measurement allows for accurate classification of activity, which can then be used to develop
activity-specific models that improve estimates of EE. The inclusion of insole pressure also
improves EE estimation within an activity classification. In particular, there was a significant
decrease in error rate in estimating cycling EE. This likely due to the changes in plantar pressure
that are associated with changes in the intensity of cycling, something difficult to detect using an
accelerometer.
It is interesting to note that we were able to achieve accurate activity
C
classification using only the 1st metatarsal pressure sensor and three-dimensional acceleration
A
(27) and the EE models also used the sensors under the metatarsals. This suggests that although
multiple sensors may be required to achieve a high level of classification and EE estimation
accuracy, it may be possible to use fewer pressure sensors without a decrease in performance. In
general, these improvements in accuracy add support to the literature demonstrating that devices
that use multiple sensors improve EE estimation (7,22,35).
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
Paragraph Number 33 The use of activity-specific branched models significantly improved
estimates of EE. In particular, RMSE decreased ~25% (0.94-0.99 to 0.69-0.77 METs) and the
width of the interval of prediction is reduced by almost 1 MET when branching was used. This
improvement is likely sensitive to classification accuracy as estimating EE on the wrong activity
could lead to substantial errors. Our classification algorithm accuracy was 100%, in part because
D
of the combination of sensors. We elected to classify 13 activities performed by the subjects in
this study into four general activity groups based on common postures and activities, rather than
TE
include more specific categories. This classification attempted to address the most common
issues encountered in EE estimation using accelerometers (underestimation of energy cost of
standing and non-weight bearing activities such as cycling) by recognizing similar activities as
C
EP
one class and using a branched model for each activity class. For example, inclusion of level,
incline/decline and loaded walking as well as running data in the “walk” class resulted in a wide
range of acceleration, pressure and metabolic values that were used to develop the walk model.
This likely improved the models ability to estimate EE during a locomotor task. Recently,
Bonomi et al. used a single hip-mounted accelerometer combined with a branched model to
classify activities and the intensity of locomotor tasks (i.e. walk and run) and reported improved
C
estimates of activity EE vs. a non-classified approach (6). Other studies have used branching
A
algorithms based on accelerometry variability (12) or heart rate and accelerometer thresholds
(9,13) but without activity classification and have also reported improved EE accuracy.
Collectively, these results support the use of branching models (with and without activity
classification) to improve EE estimation.
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
Paragraph Number 34 Developing EE prediction models based on activity classification, while
intuitively appealing, also raises important questions. Chief among them is how many different
classification groups are necessary. As noted above, we used a very general activity
classification.
While this may have improved the prediction accuracy with our modeling
approach, a narrower classification category (e.g. level walking) may allowed for the
D
development of less complex models to predict activity-specific EE. For example, Bonomi et al.
classified activity and estimated EE using a standard value from a compendium (6). Recent
TE
investigations have classified common activities based on a single accelerometer and have
elected to use more classes of activities. Bonomi et al. identified six activities (lie, sit/stand,
active standing, walk, run and cycle) while Staudenmayer et al. identified 18 activities ranging
C
EP
from washing dishes to running (6,28). If the focus of a device is to identify the time spent in
various activities, a large number of potential activities would seem important. However, a large
number of activity designations may make the combination of activity classification/EE
estimation more complex, without a marked improvement in EE prediction accuracy given the
similarity in EE form many activities. Clearly, additional research is needed to determine the
relationship between activity classes and EE prediction accuracy.
C
Paragraph Number 35 Despite very good performance of the proposed model for energy
A
expenditure prediction, a limitation of the stated results is a relatively small sample size (16
subjects). However, we introduced a wearable shoe-based system and aimed to test the overall
feasibility of EE prediction using this new device on a relatively small pilot sample. Despite its
small size the sample covers a wide range of weight/height/BMI characteristics of subjects. As
seen from the confidence intervals for RMSE for the proposed BACC-PS model even the upper
limit of the interval (0.86 METs) fits within the currently reported results from existing studies
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
on EE prediction (8,28), which allows us to conclude that the current results are reliable with the
current sample size. Future work will include collection of a significantly larger data set (with
respect to number of subjects and the variety and length of activities) to support results provided
in this paper.
Paragraph Number 36 While our results are encouraging and suggest that a footwear based
D
system can provide accurate estimates of EE, such a system is not without limitations. For
example, pressure and acceleration data outside of the laboratory may be different for a given
TE
activity (e.g. cycling) and thus present a challenge for accurate EE estimation. In addition,
footwear based systems are only viable when the individual is wearing shoes. While this is a
strength for worksite or school or other daytime monitoring, it may present a challenge if the
C
EP
objective is estimation of EE during all waking hours. Future studies that explore optimal
classification categories and test devices in a free-living (including outdoor) setting are clearly
needed. Footwear based systems will also have to be extremely rugged to withstand the
environmental and physical challenges associated with this location.
Paragraph Number 37 In summary, our results suggest that measuring the acceleration and
insole pressure in the shoe of a single foot can be used to classify activity and when combined
C
with a branched model can accurately estimate the EE associated with common daily postures
A
and activities. The accuracy and unobtrusiveness of a footwear-based device may become an
effective weight management tool.
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
ACKNOWLEDGMENTS
This work was supported in part by the University of Colorado Technology Transfer Office Proof
of Concept Grant, and NIH grant 1R43DK083229.
CONFLICT OF INTEREST
D
The results of the present study do not constitute endorsement by ACSM.
Dr. Sazonov and Dr. Browning have equity interest in Physical Activity Innovations LLC
A
C
C
EP
TE
(recipient of NIH grant 1R43DK083229 used to support this work)
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
REFERENCES
1. Ainsworth BE, Haskell WL, Leon AS, et al. Compendium of physical activities: classification
of energy costs of human physical activities. Med Sci Sports Exerc. 1993;25(1):71-80.
2. Ainsworth BE, Haskell WL, Whitt MC, et al. Compendium of physical activities: an update of
activity codes and MET intensities. Med Sci Sports Exerc. 2000;32(9 Suppl):S498-504.
D
3. Bamberg SJM, Benbasat AY, Scarborough DM, Krebs DE, Paradiso JA. Gait analysis using a
shoe-integrated wireless sensor system. IEEE Trans Inf Technol Biomed. 2008;12(4):413-423.
TE
4. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of
clinical measurement. Lancet. 1986;1(8476):307-310.
5. Bonomi AG, Plasqui G, Goris AH, Westerterp, K R. Estimation of Free-Living Energy
C
EP
Expenditure Using a Novel Activity Monitor Designed to Minimize Obtrusiveness. Obesity
(Silver Spring). 2010;18(9):1845-1851.
6. Bonomi AG, Plasqui G, Goris AHC, Westerterp KR. Improving assessment of daily energy
expenditure by identifying types of physical activity with a single accelerometer. J Appl Physiol.
2009;107(3):655-661.
7. Brage S, Brage N, Franks PW, Ekelund U, Wareham NJ. Reliability and validity of the
C
combined heart rate and movement sensor Actiheart. Eur J Clin Nutr. 2005;59(4):561-570.
A
8. Brage S, Ekelund U, Brage N, et al. Hierarchy of individual calibration levels for heart rate
and accelerometry to measure physical activity. J Appl Physiol. 2007;103(2):682-692.
9. Brage S, Brage N, Franks PW, et al. Branched equation modeling of simultaneous
accelerometry and heart rate monitoring improves estimate of directly measured physical activity
energy expenditure. J. Appl. Physiol. 2004;96(1):343-351.
10. Chen KY, Sun M. Improving energy expenditure estimation by using a triaxial accelerometer.
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
J Appl Physiol. 1997;83(6):2112-2122.
11. Choi L, Chen KY, Acra SA, Buchowski MS. Distributed lag and spline modeling for
predicting energy expenditure from accelerometry in youth. J Appl Physiol. 2010;108(2):314327.
predict energy expenditure. J Appl Physiol. 2006;100(4):1324-1331.
D
12. Crouter SE, Clowers KG, Bassett DR. A novel method for using accelerometer data to
13. Edwards AG, Hill JO, Byrnes WC, Browning RC. Accuracy of Optimized Branched
TE
Algorithms to Assess Activity-Specific Physical Activity Energy Expenditure. Medicine &
Science in Sports & Exercise. 2010;42(4):672-682.
14. F-Scan. Tekscan Products for Pressure Mapping and Force Measurement. Available at:
C
EP
http://www.tekscan.com/medical/system-fscan1.html [Accessed November 15, 2010].
15. Garcia AW, Langenthal CR, Angulo-Barroso RM, Gross MM. A Comparison of
Accelerometers for Predicting Energy Expenditure and Vertical Ground Reaction Force in
School-Age Children. Measurement in Physical Education and Exercise Science. 2004;8(3):119.
16. Haskell WL, Lee I, Pate RR, et al. Physical activity and public health: updated
recommendation for adults from the American College of Sports Medicine and the American
C
Heart Association. Med Sci Sports Exerc. 2007;39(8):1423-1434.
A
17. Havinga PJM, Marin-Perianu M, Thalen JP. SensorShoe: Mobile Gait Analysis for
Parkinson's Disease Patients. 2007. Technical Report TR-CTIT-07-63, Centre for Telematics and
Information Technology University of Twente, Enschede. 58 p.
18. Hendelman D, Miller K, Baggett C, Debold E, Freedson P. Validity of accelerometry for the
assessment of moderate intensity physical activity in the field. Med Sci Sports Exerc. 2000;32(9
Suppl):S442-449.
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
19. Hustvedt BE, Svendsen M, Lovo A, et al. Validation of ActiReg to measure physical activity
and energy expenditure against doubly labelled water in obese persons. Br J Nutr.
2008;100(1):219-226.
20. Jagos H, Oberzaucher J. Development of a Wearable Measurement System to Identify
Characteristics in Human Gait - eSHOE -. In: Computers Helping People with Special Needs.;
D
2008:1301-1304. Available at: http://dx.doi.org/10.1007/978-3-540-70540-6_194 [Accessed
December 16, 2009].
TE
21. Krishnamurthy V, Fowler K, Sazonov E. The effect of time synchronization of wireless
sensors on the modal analysis of structures. Smart Materials and Structures. 2008;17(5):055018.
22. Levine J, Melanson EL, Westerterp KR, Hill JO. Measurement of the components of
C
EP
nonexercise activity thermogenesis. Am J Physiol Endocrinol Metab. 2001;281(4):E670-5.
23. Padoan A. MATLAB Central - File detail - Passing and Bablok regression. Available at:
http://www.mathworks.com/matlabcentral/fileexchange/24894-passing-and-bablok-regression
[Accessed March 2, 2010].
24. Passing H, Bablok. A new biometrical procedure for testing the equality of measurements
from two different analytical methods. Application of linear regression procedures for method
A
720.
C
comparison studies in clinical chemistry, Part I. J. Clin. Chem. Clin. Biochem. 1983;21(11):709-
25. Plasqui G, Westerterp KR. Physical Activity Assessment With Accelerometers: An
Evaluation Against Doubly Labeled Water[ast][ast]. Obesity. 2007;15(10):2371-2379.
26. Sazonov E, Fulk G, Yves S, Hill J, Browning R. Monitoring of posture allocations and
activities by a shoe-based wearable sensor. IEEE Transactions on Bio-Medical Engineering,
accepted. 2010.
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
27. Sazonov ES, Fulk G, Sazonova N, Schuckers S. Automatic Recognition of postures and
activities in stroke patients. Conf Proc IEEE Eng Med Biol Soc. 2009;1:2200-2203.
28. Staudenmayer J, Pober D, Crouter SE, Bassett DR, Freedson P. An artificial neural network
to estimate physical activity energy expenditure and identify physical activity type from an
accelerometer. J Appl Physiol. 2009; 107: 1300-1307.
D
29. Strath SJ, Bassett DR, Swartz AM, Thompson DL. Simultaneous heart rate-motion sensor
technique to estimate energy expenditure. Med Sci Sports Exerc. 2001;33(12):2118-2123.
TE
30. Strath SJ, Bassett DR, Thompson DL, Swartz AM. Validity of the simultaneous heart ratemotion sensor technique for measuring energy expenditure. Med Sci Sports Exerc.
2002;34(5):888-894.
C
EP
31. Systems pedar. Available at: http://novel.de/novelcontent/pedar [Accessed November 15,
2010].
32. Tharion WJ, Yokota M, Buller MJ, DeLany JP, Hoyt RW. Total energy expenditure estimated
using a foot-contact pedometer. Med. Sci. Monit. 2004;10(9):CR504-509.
33.United States Department of Health and Human Services. The Surgeon General's Call to
Action to Prevent and Decrease Overweight and Obesity. Rockville, Md.: Public Health Service,
C
Office of the Surgeon General, 2001. 39 p. Available from: U.S. GPO, Washington.
A
34. Zhang K, Pi-Sunyer FX, Boozer CN. Improving energy expenditure estimation for physical
activity. Med Sci Sports Exerc. 2004;36(5):883-889.
35. Zhang K, Werner P, Sun M, Pi-Sunyer FX, Boozer CN. Measurement of Human Daily
Physical Activity. Obesity. 2003;11(1): 33-40.
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
FIGURE CAPTIONS
Fig. 1. Shoe device: (a) Overall view of the shoe device; (b) The rear view of a shoe including
the accelerometer, battery and power switch; (c) Pressure-sensitive insole with 5 pressure
sensors: heel (1), 3rd metatarsal head (2), 1st metatarsal head (3), 5th metatarsal head (4), hallux
(5).
D
Fig. 2. Bland-Altman plots for shoe-based models: (a) BACC-PS model, kcal·min-1, (b) BACC
kcal·min-1.
TE
model, kcal·min-1, (c) non-branched ACC-PS model, kcal·min-1, (d) non-branched ACC model,
Fig. 3. Passing-Bablok regression plots for shoe-based models: (a) BACC-PS model, kcal·min -1,
(b) BACC model, kcal·min-1; (c) non-branched ACC-PS model, kcal·min-1, (d) non-branched
A
C
C
EP
ACC model, kcal·min-1.
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
TABLES
Table 1. Subject characteristics
Table 2. Study protocol
Table 3. Energy expenditure prediction by minute
D
SUPPLEMENTAL DIGITAL CONTENT FILES
accelerometer; (b) 5 pressure sensors.
TE
SDC1.tif: Figure. Sample raw sensors signal for “walking 2.5 mph” activity: (a) 3-dim
SDC2.pdf: Table . Regression coefficients for the best BACC-PS model (EE in kcal/min)
SDC3.pdf: Table. Regression coefficients for the best BACC model (EE in kcal/min)
C
EP
SDC4.pdf: Table. Regression coefficients for the best non-branched ACC-PS and ACC models
SDC5.pdf: Table. Examination of the presence of fixed and proportional bias and linearity
SDC6.pdf: Table. Comparison of BACC-PS model performance using predictors from single
A
C
shoe and both shoes
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
A
C
C
EP
TE
D
Figure 1
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
A
C
C
EP
TE
D
Figure 2
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
A
C
C
EP
TE
D
Figure 3
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
Table 1. Subject characteristics
Women (N=8)
Mean ± SD
Range
66.9 ± 16.8
48.6-100.9
69.3 ± 1.8
67.0-72.0
64.3 ± 2.8
BMI, kg∙m-2
28.0 ± 5.9
18.9-35.8
25.4 ± 7.3
Age, yr
25.6 ± 8.6
18-44
24.4 ± 3.9
Shoe size, US
10.3 ± 0.6
9.5-11.0
TE
Height, in.
18.1-39.4
18-29
7.0-9.0
A
C
C
EP
7.9 ± 0.7
61.0-70.0
D
Weight, kg
Men (N=8)
Mean ± SD
Range
86.8 ± 20.0
59.0-119.8
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
Table 2. Study protocol
Description
Sit quietly
Posture/Activity Group
Sit
2
Stand quietly
Level Treadmill Walking/Jogging
Stand
3
0.67 m/s (1.5 mph)
Walk/Jog
4
1.11 m/s (2.5 mph)
Walk/Jog
5
1.56 m/s (3.5 mph)
Walk/Jog
6
7
8
2.00 m/s (4.5 mph) - jogging
Ascend/Descend stairs*
Sit with fidgeting
9
Stand with fidgeting
Treadmill Walking
10
1.11 m/s +1.5% grade
Walk/Jog
11
1.11 m/s -1.5% grade
Walk/Jog
12
1.11 m/s with 10% of body weight
Walk/Jog
TE
D
Trial
1
Walk/Jog
Sit
C
EP
Stand
13
C
held in bags (5% held by each hand)
Cycling:
50W, 50 rpm
Cycle
A
14
100W, 75rpm
Cycle
* Metabolic data not collected during stair ascent/descent
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
Table 3. Energy expenditure prediction by minute
Number of
Branch
95% CI for
Model
1-min
RMSEMET
Bias, METs
RMSEMET
agreement, METs
0.2593
0.3186
0.7550
0.96633
0.6870
0.2593
0.3285
0.7544
1.295
0.7679
(0.14, 0.37)
(0.20, 0.42)
(0.55, 0.96)
(0.60, 1.34)
(0.53, 0.86)
(0.14, 0.37)
(0.18, 0.49)
(0.53, 1.00)
(0.92, 1.70)
(0.58, 0.98)
3.0
5.8
3.6
3.4
3.85
3.0
5.6
3.6
8.9
4.7
ACC-PS non-branched
ACC non-branched
197
197
0.9389
0.9886
(0.78, 1.15)
(0.76, 1.22)
3.1
2.7
0.0276
0.0323
0.0466
0.0617
0.0437
0.0276
0.0408
0.0385
0.1517
0.0550
(-0.50, 0.55)
(-0.61, 0.67)
(-1.47, 1.56)
(-1.90,2.02)
(-1.33, 1.42)
(-0.50, 0.55)
(-0.62, 0.70)
(-1.48, 1.55)
(-2.46, 2.77)
(-1.48, 1.59)
0.0395
0.0459
(-1.84, 1.92)
(-1.93, 2.03)
TE
BACC-PS Sit
Stand
Walk
Cycle
Aggregated
BACC
Sit
Stand
Walk
Cycle
Aggregated
recordings
31
32
103
31
197
31
32
103
31
197
D
model
95% interval of
ARD, %
A
C
C
EP
BACC-PS model uses branching into 4 sub-models corresponding to sitting, standing, walking and cycling activities; it also uses signals from
both accelerometer and pressure sensors for prediction of EE.
BACC is a branched model that only uses accelerometer measures for prediction of EE.
ACC-PS is a non-branched model that uses signals from both accelerometer and pressure sensors for prediction of EE.
ACC is a non-branched model that only uses accelerometer measures for prediction of EE.
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
A
C
C
EP
TE
D
Supplemental Digital Content Fig.1
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
A
C
C
EP
TE
D
Supplemental Digital Content Fig.2
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
A
C
C
EP
TE
D
Supplemental Digital Content Fig.3
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
A
C
C
EP
TE
D
Supplemental Digital Content Fig.4
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
A
C
C
EP
TE
D
Supplemental Digital Content Fig.5
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
A
C
C
EP
TE
D
Supplemental Digital Content Fig.6
Copyright © 2010 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
View publication stats