Estimating metabolic equivalents for activities in daily life using acceleration and heart rate in wearable devices

Motofumi Nakanishi, Shintaro Izumi, Sho Nagayoshi, Hiroshi Kawaguchi, Masahiko Yoshimoto, Toshikazu Shiga, Takafumi Ando, Satoshi Nakae, Chiyoko Usui, Tomoko Aoyama, Shigeho Tanaka

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Background: Herein, an algorithm that can be used in wearable health monitoring devices to estimate metabolic equivalents (METs) based on physical activity intensity data, particularly for certain activities in daily life that make MET estimation difficult. Results: Energy expenditure data were obtained from 42 volunteers using indirect calorimetry, triaxial accelerations and heart rates. The proposed algorithm used the percentage of heart rate reserve (%HRR) and the acceleration signal from the wearable device to divide the data into a middle-intensity group and a high-intensity group (HIG). The two groups were defined in terms of estimated METs. Evaluation results revealed that the classification accuracy for both groups was higher than 91%. To further facilitate MET estimation, five multiple-regression models using different features were evaluated via leave-one-out cross-validation. Using this approach, all models showed significant improvements in mean absolute percentage error (MAPE) of METs in the HIG, which included stair ascent, and the maximum reduction in MAPE for HIG was 24% compared to the previous model (HJA-750), which demonstrated a 70.7% improvement ratio. The most suitable model for our purpose that utilized heart rate and filtered synthetic acceleration was selected and its estimation error trend was confirmed. Conclusion: For HIG, the MAPE recalculated by the most suitable model was 10.5%. The improvement ratio was 71.6% as compared to the previous model (HJA-750C). This result was almost identical to that obtained from leave-one-out cross-validation. This proposed algorithm revealed an improvement in estimation accuracy for activities in daily life; in particular, the results included estimated values associated with stair ascent, which has been a difficult activity to evaluate so far.

Original languageEnglish
Article number100
JournalBioMedical Engineering Online
Volume17
Issue number1
DOIs
Publication statusPublished - 2018 Jul 28
Externally publishedYes

Fingerprint

Metabolic Equivalent
Heart Rate
Equipment and Supplies
Stairs
Indirect Calorimetry
Energy Metabolism
Calorimetry
Volunteers
Error analysis
Exercise
Health
Monitoring

Keywords

  • Energy expenditure estimations
  • Heart rate
  • Metabolic equivalents
  • Physical activity
  • Physical activity classification
  • Triaxial acceleration

ASJC Scopus subject areas

  • Radiological and Ultrasound Technology
  • Biomaterials
  • Biomedical Engineering
  • Radiology Nuclear Medicine and imaging

Cite this

Estimating metabolic equivalents for activities in daily life using acceleration and heart rate in wearable devices. / Nakanishi, Motofumi; Izumi, Shintaro; Nagayoshi, Sho; Kawaguchi, Hiroshi; Yoshimoto, Masahiko; Shiga, Toshikazu; Ando, Takafumi; Nakae, Satoshi; Usui, Chiyoko; Aoyama, Tomoko; Tanaka, Shigeho.

In: BioMedical Engineering Online, Vol. 17, No. 1, 100, 28.07.2018.

Research output: Contribution to journalArticle

Nakanishi, M, Izumi, S, Nagayoshi, S, Kawaguchi, H, Yoshimoto, M, Shiga, T, Ando, T, Nakae, S, Usui, C, Aoyama, T & Tanaka, S 2018, 'Estimating metabolic equivalents for activities in daily life using acceleration and heart rate in wearable devices', BioMedical Engineering Online, vol. 17, no. 1, 100. https://doi.org/10.1186/s12938-018-0532-2
Nakanishi, Motofumi ; Izumi, Shintaro ; Nagayoshi, Sho ; Kawaguchi, Hiroshi ; Yoshimoto, Masahiko ; Shiga, Toshikazu ; Ando, Takafumi ; Nakae, Satoshi ; Usui, Chiyoko ; Aoyama, Tomoko ; Tanaka, Shigeho. / Estimating metabolic equivalents for activities in daily life using acceleration and heart rate in wearable devices. In: BioMedical Engineering Online. 2018 ; Vol. 17, No. 1.
@article{ca0b616e3ff74ae1a1ff286547110a5b,
title = "Estimating metabolic equivalents for activities in daily life using acceleration and heart rate in wearable devices",
abstract = "Background: Herein, an algorithm that can be used in wearable health monitoring devices to estimate metabolic equivalents (METs) based on physical activity intensity data, particularly for certain activities in daily life that make MET estimation difficult. Results: Energy expenditure data were obtained from 42 volunteers using indirect calorimetry, triaxial accelerations and heart rates. The proposed algorithm used the percentage of heart rate reserve ({\%}HRR) and the acceleration signal from the wearable device to divide the data into a middle-intensity group and a high-intensity group (HIG). The two groups were defined in terms of estimated METs. Evaluation results revealed that the classification accuracy for both groups was higher than 91{\%}. To further facilitate MET estimation, five multiple-regression models using different features were evaluated via leave-one-out cross-validation. Using this approach, all models showed significant improvements in mean absolute percentage error (MAPE) of METs in the HIG, which included stair ascent, and the maximum reduction in MAPE for HIG was 24{\%} compared to the previous model (HJA-750), which demonstrated a 70.7{\%} improvement ratio. The most suitable model for our purpose that utilized heart rate and filtered synthetic acceleration was selected and its estimation error trend was confirmed. Conclusion: For HIG, the MAPE recalculated by the most suitable model was 10.5{\%}. The improvement ratio was 71.6{\%} as compared to the previous model (HJA-750C). This result was almost identical to that obtained from leave-one-out cross-validation. This proposed algorithm revealed an improvement in estimation accuracy for activities in daily life; in particular, the results included estimated values associated with stair ascent, which has been a difficult activity to evaluate so far.",
keywords = "Energy expenditure estimations, Heart rate, Metabolic equivalents, Physical activity, Physical activity classification, Triaxial acceleration",
author = "Motofumi Nakanishi and Shintaro Izumi and Sho Nagayoshi and Hiroshi Kawaguchi and Masahiko Yoshimoto and Toshikazu Shiga and Takafumi Ando and Satoshi Nakae and Chiyoko Usui and Tomoko Aoyama and Shigeho Tanaka",
year = "2018",
month = "7",
day = "28",
doi = "10.1186/s12938-018-0532-2",
language = "English",
volume = "17",
journal = "BioMedical Engineering Online",
issn = "1475-925X",
publisher = "BioMed Central",
number = "1",

}

TY - JOUR

T1 - Estimating metabolic equivalents for activities in daily life using acceleration and heart rate in wearable devices

AU - Nakanishi, Motofumi

AU - Izumi, Shintaro

AU - Nagayoshi, Sho

AU - Kawaguchi, Hiroshi

AU - Yoshimoto, Masahiko

AU - Shiga, Toshikazu

AU - Ando, Takafumi

AU - Nakae, Satoshi

AU - Usui, Chiyoko

AU - Aoyama, Tomoko

AU - Tanaka, Shigeho

PY - 2018/7/28

Y1 - 2018/7/28

N2 - Background: Herein, an algorithm that can be used in wearable health monitoring devices to estimate metabolic equivalents (METs) based on physical activity intensity data, particularly for certain activities in daily life that make MET estimation difficult. Results: Energy expenditure data were obtained from 42 volunteers using indirect calorimetry, triaxial accelerations and heart rates. The proposed algorithm used the percentage of heart rate reserve (%HRR) and the acceleration signal from the wearable device to divide the data into a middle-intensity group and a high-intensity group (HIG). The two groups were defined in terms of estimated METs. Evaluation results revealed that the classification accuracy for both groups was higher than 91%. To further facilitate MET estimation, five multiple-regression models using different features were evaluated via leave-one-out cross-validation. Using this approach, all models showed significant improvements in mean absolute percentage error (MAPE) of METs in the HIG, which included stair ascent, and the maximum reduction in MAPE for HIG was 24% compared to the previous model (HJA-750), which demonstrated a 70.7% improvement ratio. The most suitable model for our purpose that utilized heart rate and filtered synthetic acceleration was selected and its estimation error trend was confirmed. Conclusion: For HIG, the MAPE recalculated by the most suitable model was 10.5%. The improvement ratio was 71.6% as compared to the previous model (HJA-750C). This result was almost identical to that obtained from leave-one-out cross-validation. This proposed algorithm revealed an improvement in estimation accuracy for activities in daily life; in particular, the results included estimated values associated with stair ascent, which has been a difficult activity to evaluate so far.

AB - Background: Herein, an algorithm that can be used in wearable health monitoring devices to estimate metabolic equivalents (METs) based on physical activity intensity data, particularly for certain activities in daily life that make MET estimation difficult. Results: Energy expenditure data were obtained from 42 volunteers using indirect calorimetry, triaxial accelerations and heart rates. The proposed algorithm used the percentage of heart rate reserve (%HRR) and the acceleration signal from the wearable device to divide the data into a middle-intensity group and a high-intensity group (HIG). The two groups were defined in terms of estimated METs. Evaluation results revealed that the classification accuracy for both groups was higher than 91%. To further facilitate MET estimation, five multiple-regression models using different features were evaluated via leave-one-out cross-validation. Using this approach, all models showed significant improvements in mean absolute percentage error (MAPE) of METs in the HIG, which included stair ascent, and the maximum reduction in MAPE for HIG was 24% compared to the previous model (HJA-750), which demonstrated a 70.7% improvement ratio. The most suitable model for our purpose that utilized heart rate and filtered synthetic acceleration was selected and its estimation error trend was confirmed. Conclusion: For HIG, the MAPE recalculated by the most suitable model was 10.5%. The improvement ratio was 71.6% as compared to the previous model (HJA-750C). This result was almost identical to that obtained from leave-one-out cross-validation. This proposed algorithm revealed an improvement in estimation accuracy for activities in daily life; in particular, the results included estimated values associated with stair ascent, which has been a difficult activity to evaluate so far.

KW - Energy expenditure estimations

KW - Heart rate

KW - Metabolic equivalents

KW - Physical activity

KW - Physical activity classification

KW - Triaxial acceleration

UR - http://www.scopus.com/inward/record.url?scp=85050664504&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85050664504&partnerID=8YFLogxK

U2 - 10.1186/s12938-018-0532-2

DO - 10.1186/s12938-018-0532-2

M3 - Article

C2 - 30055617

AN - SCOPUS:85050664504

VL - 17

JO - BioMedical Engineering Online

JF - BioMedical Engineering Online

SN - 1475-925X

IS - 1

M1 - 100

ER -