The objective of this study was to investigate the accuracy of a Deep Neural Network (DNN) in recognizing activities typical for hospitalized patients. A data collection study was conducted with 20 ...healthy volunteers (10 males and 10 females, age = 43 ± 13 years) in a simulated hospital environment. A single triaxial accelerometer mounted on the trunk was used to measure body movement and recognize six activity types: lying in bed, upright posture, walking, wheelchair transport, stair ascent and stair descent. A DNN consisting of a three-layer convolutional neural network followed by a long short-term memory layer was developed for this classification problem. Additionally, features were extracted from the accelerometer data to train a support vector machine (SVM) classifier for comparison. The DNN reached 94.52% overall accuracy on the holdout dataset compared to 83.35% of the SVM classifier. In conclusion, a DNN is capable of recognizing types of physical activity in simulated hospital conditions using data captured by a single tri-axial accelerometer. The method described may be used for continuous monitoring of patient activities during hospitalization to provide additional insights into the recovery process.
Obesity represents a strong risk factor for developing chronic diseases. Strategies for disease prevention often promote lifestyle changes encouraging participation in physical activity. However, ...determining what amount of physical activity is necessary for achieving specific health benefits has been hampered by the lack of accurate instruments for monitoring physical activity and the related physiological outcomes. This review aims at presenting recent advances in activity-monitoring technology and their application to support interventions for health promotion. Activity monitors have evolved from step counters and measuring devices of physical activity duration and intensity to more advanced systems providing quantitative and qualitative information on the individuals' activity behavior. Correspondingly, methods to predict activity-related energy expenditure using bodily acceleration and subjects characteristics have advanced from linear regression to innovative algorithms capable of determining physical activity types and the related metabolic costs. These novel techniques can monitor modes of sedentary behavior as well as the engagement in specific activity types that helps to evaluate the effectiveness of lifestyle interventions. In conclusion, advances in activity monitoring have the potential to support the design of response-dependent physical activity recommendations that are needed to generate effective and personalized lifestyle interventions for health promotion.
Accurate identification of physical activity types has been achieved in laboratory conditions using single-site accelerometers and classification algorithms. This methodology is then applied to ...free-living subjects to determine activity behavior. This study is aimed at analyzing the reproducibility of the accuracy of laboratory-trained classification algorithms in free-living subjects during daily life. A support vector machine (SVM), a feed-forward neural network (NN), and a decision tree (DT) were trained with data collected by a waist-mounted accelerometer during a laboratory trial. The reproducibility of the classification performance was tested on data collected in daily life using a multiple-site accelerometer augmented with an activity diary for 20 healthy subjects (age: 30 ± 9; BMI: 23.0 ± 2.6 kg/m 2 ). Leave-one-subject-out cross validation of the training data showed accuracies of 95.1 ± 4.3%, 91.4 ± 6.7%, and 92.2 ± 6.6% for the SVM, NN, and DT, respectively. All algorithms showed a significantly decreased accuracy in daily life as compared to the reference truth represented by the IDEEA and diary classifications (75.6 ± 10.4%, 74.8 ± 9.7%, and 72.2 ± 10.3 %; p <; 0.05). In conclusion, cross validation of training data overestimates the accuracy of the classification algorithms in daily life.
This study aimed to model below and above anaerobic threshold exercise-induced heart rate (HR) drift, so that the corrected HR could better represent
kinetics during and after the exercise itself.
...Fifteen healthy subjects (age: 28 ± 5 years;
: 50 ± 8 mL/kg/min; 5 females) underwent a maximal and a 30-min submaximal (80% of the anaerobic threshold) running exercises. A five-stage computational (i.e., delay block, new training impulse-calculation block, Sigmoid correction block, increase block, and decrease block) model was built to account for instantaneous HR, fitness, and age and to onset, increase, and decrease according to the exercise intensity and duration.
The area under the curve (AUC) of the hysteresis function, which described the differences in the maximal and submaximal exercise-induced
and HR kinetics, was significantly reduced for both maximal (26%) and submaximal (77%) exercises and consequent recoveries.
In conclusion, this model allowed HR drift instantaneous correction, which could be exploited in the future for more accurate
estimations.
The aim of this study was to develop models for the detection of type, duration, and intensity of human physical activity using one triaxial accelerometer.
Twenty subjects (age = 29 +/- 6 yr, BMI = ...23.6 +/- 3.2 kg.m) performed 20 selected activities, including walking, running, and cycling, wearing one triaxial accelerometer mounted on the lower back. Identification of activity type was based on a decision tree. The decision tree evaluated attributes (features) of the acceleration signal. The features were measured in intervals of defined duration (segments). Segment size determined the time resolution of the decision tree to assess activity duration. Decision trees with a time resolution of 0.4, 0.8, 1.6, 3.2, 6.4, and 12.8 s were developed, and the respective classification performances were evaluated. Multiple linear regression was used to estimate speed of walking, running, and cycling based on acceleration features.
Maximal accuracy for the classification of activity type (93%) was reached when the segment size of analysis was 6.4 or 12.8 s. The smaller the segment size, the lower the classification accuracy achieved. Segments of 6.4 s gave the highest time resolution for measuring activity duration without decreasing the classification accuracy. The developed models estimated walking, running, and cycling speeds with a standard error of 0.20, 1.26, and 1.36 km.h, respectively.
This study demonstrated the ability of a triaxial accelerometer in detecting type, duration, and intensity of physical activity using models based on acceleration features. Future studies are needed to validate the presented models in free-living conditions.
Muography of different structures using muon scattering and absorption algorithms Vanini, S; Calvini, P; Checchia, P ...
Philosophical transactions - Royal Society. Mathematical, Physical and engineering sciences/Philosophical transactions - Royal Society. Mathematical, physical and engineering sciences,
12/2018, Letnik:
377, Številka:
2137
Journal Article
Recenzirano
Odprti dostop
In recent decades, muon imaging has found a plethora of applications in many fields. This technique succeeds to infer the density distribution of big inaccessible structures where conventional ...techniques cannot be used. The requirements of different applications demand specific implementations of image reconstruction algorithms for either multiple scattering or absorption-transmission data analysis, as well as noise-suppression filters and muon momentum estimators. This paper presents successful results of image reconstruction techniques applied to simulated data of some representative applications. In addition to well-known reconstruction methods, a novel approach, the so-called μCT, is proposed for the inspection of spent nuclear fuel canisters. Results obtained based on both μCT and the maximum-likelihood expectation maximization reconstruction algorithms are presented.This article is part of the Theo Murphy meeting issue 'Cosmic-ray muography'.
1 Department of Human Biology, Maastricht University, Maastricht; and 2 Group Care and Health Applications and 3 DirectLife New Wellness Solutions, Philips Research Laboratories, Eindhoven, The ...Netherlands
Submitted 11 February 2009
; accepted in final form 23 June 2009
Accelerometers are often used to quantify the acceleration of the body in arbitrary units (counts) to measure physical activity (PA) and to estimate energy expenditure. The present study investigated whether the identification of types of PA with one accelerometer could improve the estimation of energy expenditure compared with activity counts. Total energy expenditure (TEE) of 15 subjects was measured with the use of double-labeled water. The physical activity level (PAL) was derived by dividing TEE by sleeping metabolic rate. Simultaneously, PA was measured with one accelerometer. Accelerometer output was processed to calculate activity counts per day (AC D ) and to determine the daily duration of six types of common activities identified with a classification tree model. A daily metabolic value (MET D ) was calculated as mean of the MET compendium value of each activity type weighed by the daily duration. TEE was predicted by AC D and body weight and by AC D and fat-free mass, with a standard error of estimate (SEE) of 1.47 MJ/day, and 1.2 MJ/day, respectively. The replacement in these models of AC D with MET D increased the explained variation in TEE by 9%, decreasing SEE by 0.14 MJ/day and 0.18 MJ/day, respectively. The correlation between PAL and MET D ( R 2 = 51%) was higher than that between PAL and AC D ( R 2 = 46%). We conclude that identification of activity types combined with MET intensity values improves the assessment of energy expenditure compared with activity counts. Future studies could develop models to objectively assess activity type and intensity to further increase accuracy of the energy expenditure estimation.
double-labeled water; motion sensor; classification tree; activity recognition
Address for reprint requests and other correspondence: A. G. Bonomi, Human Biology, Maastricht Univ., PO Box 616, 6200 MD Maastricht, The Netherlands (e-mail: a.bonomi{at}HB.unimaas.nl )
Low cardiorespiratory fitness (CRF) increases risk of all-cause mortality and cardiovascular events. Periodic CRF assessment can have an important preventive function. The objective of this study was ...to develop a protocol-free method to estimate CRF in daily life based on heart rate (HR) and body acceleration measurements. Acceleration and HR data were collected from 37 subjects (men = 49%) while they performed a standardized laboratory activity protocol (sitting, walking, running, cycling) and during a 5-day free-living monitoring period. CRF was determined by oxygen uptake (V̇o
) during maximal exercise testing. A doubly labeled water-validated equation was used to predict total energy expenditure (TEE) from acceleration data. A fitness index was defined as the ratio between TEE and HR (TEE-pulse). Activity recognition techniques were used to process acceleration features and classify sedentary, ambulatory, and other activity types. Regression equations based on TEE-pulse data from each activity type were developed to predict V̇o
. TEE-pulse measured within each activity type of the laboratory protocol was highly correlated with V̇o
(
from 0.74-0.91). Averaging the outcome of each activity-type specific equation based on TEE-pulse from the laboratory data led to accurate estimates of V̇o
root mean square error (RMSE): 300 mL O
/min, or 10%. The difference between laboratory and free-living determined TEE-pulse was 3.7 ± 11% (
= 0.85). The prediction method preserved the prediction accuracy when applied to free-living data (RMSE: 367 mL O
/min, or 12%). Measurements of body acceleration and HR can be used to predict V̇o
in daily life. Activity-specific prediction equations are needed to achieve highly accurate estimates of CRF.
This is among the very few studies validating, in free-living conditions, a method to estimate cardiorespiratory fitness using heart rate and body acceleration data. A novel parameter called TEE-pulse, which was defined as the ratio between accelerometer-determined energy expenditure and heart rate, was highly correlated with maximal oxygen uptake (V̇o
). Activity classification and the use of activity-selective prediction equations outperformed previously published methods for estimating V̇o
from heart rate and acceleration data.
Activity energy expenditure (AEE) is the component of daily energy expenditure that is mainly influenced by the amount of physical activity (PA) and by the weight of the body displaced. This study ...aimed at analyzing the effect of weight loss on PA and AEE. The body weight and PA of 66 overweight and obese subjects were measured at baseline and after 12 weeks of 67% energy restriction. PA was measured using a tri-axial accelerometer for movement registration (Tracmor) and quantified in activity counts. Tracmor recordings were also processed using a classification algorithm to recognize 6 common activity types engaged in during the day. A doubly-labeled water validated equation based on Tracmor output was used to estimate AEE. After weight loss, body weight decreased by 13±4%, daily activity counts augmented by 9% (95% CI: +2%, +15%), and this increase was weakly associated with the decrease in body weight (R(2) = 7%; P<0.05). After weight loss subjects were significantly (P<0.05) less sedentary (-26 min/d), and increased the time spent walking (+11 min/d) and bicycling (+4 min/d). However, AEE decreased by 0.6±0.4 MJ/d after weight loss. On average, a 2-hour/day reduction of sedentary time by increasing ambulatory and generic activities was required to restore baseline levels of AEE. In conclusion, after weight loss PA increased but the related metabolic demand did not offset the reduction in AEE due to the lower body weight. Promoting physical activity according to the extent of weight loss might increase successfulness of weight maintenance.
Celotno besedilo
Dostopno za:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK
Cardio-respiratory fitness (CRF) is a widespread essential indicator in Sports Science as well as in Sports Medicine. This study aimed to develop and validate a prediction model for CRF based on a 45 ...second self-test, which can be conducted anywhere. Criterion validity, test re-test study was set up to accomplish our objectives. Data from 81 healthy volunteers (age: 29 ± 8 years, BMI: 24.0 ± 2.9), 18 of whom females, were used to validate this test against gold standard. Nineteen volunteers repeated this test twice in order to evaluate its repeatability. CRF estimation models were developed using heart rate (HR) features extracted from the resting, exercise, and the recovery phase. The most predictive HR feature was the intercept of the linear equation fitting the HR values during the recovery phase normalized for the height2 (r2 = 0.30). The Ruffier-Dickson Index (RDI), which was originally developed for this squat test, showed a negative significant correlation with CRF (r = -0.40), but explained only 15% of the variability in CRF. A multivariate model based on RDI and sex, age and height increased the explained variability up to 53% with a cross validation (CV) error of 0.532 L ∙ min-1 and substantial repeatability (ICC = 0.91). The best predictive multivariate model made use of the linear intercept of HR at the beginning of the recovery normalized for height2 and age2; this had an adjusted r2 = 0. 59, a CV error of 0.495 L·min-1 and substantial repeatability (ICC = 0.93). It also had a higher agreement in classifying CRF levels (κ = 0.42) than RDI-based model (κ = 0.29). In conclusion, this simple 45 s self-test can be used to estimate and classify CRF in healthy individuals with moderate accuracy and large repeatability when HR recovery features are included.
Celotno besedilo
Dostopno za:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK