Ambulatory system for human motion analysis using a kinematic sensor

Monitoring of daily physical activity in the elderly

Bijan Najafi, Kamiar Aminian, Anisoara Paraschiv-Ionescu, François Loew, Christophe J. Büla, Philippe Robert

Research output: Contribution to journalArticle

533 Citations (Scopus)

Abstract

A new method of physical activity monitoring is presented, which is able to detect body postures (sitting, standing, and lying) and periods of walking in elderly persons using only one kinematic sensor attached to the chest. The wavelet transform, in conjunction with a simple kinematics model, was used to detect different postural transitions (PTs) and walking periods during daily physical activity. To evaluate the system, three studies were performed. The method was first tested on 11 community-dwelling elderly subjects in a gait laboratory where an optical motion system (Vicon) was used as a reference system. In the second study, the system was tested for classifying PTs (i.e., lying-to-sitting, sitting-to-lying, and turning the body in bed) in 24 hospitalized elderly persons. Finally, in a third study monitoring was performed on nine elderly persons for 45-60 min during their daily physical activity. Moreover, the possibility-to-perform long-term monitoring over 12 h has been shown. The first study revealed a close concordance between the ambulatory and reference systems. Overall, subjects performed 349 PTs during this study. Compared with the reference system, the ambulatory system had an overall sensitivity of 99% for detection of the different PTs. Sensitivities and specificities were 93% and 82% in sit-to-stand, and 82% and 94% in stand-to-sit, respectively. In both first and second studies, the ambulatory system also showed a very high accuracy (> 99%) in identifying the 62 transfers or rolling out of bed, as well as 144 different posture changes to the back, ventral, right and left sides. Relatively high sensitivity (> 90%) was obtained for the classification of usual physical activities in the third study in comparison with visual observation. Sensitivities and specificities were, respectively, 90.2% and 93.4% in sitting, 92.2% and 92.1% in "standing + walking," and, finally, 98.4% and 99.7% in lying. Overall detection errors (as percent of range) were 3.9% for "standing + walking," 4.1% for sitting, and 0.3% for lying. Finally, overall symmetric mean average errors were 12% for "standing + walking," 8.2% for sitting, and 1.3% for lying.

Original languageEnglish (US)
Pages (from-to)711-723
Number of pages13
JournalIEEE Transactions on Biomedical Engineering
Volume50
Issue number6
DOIs
StatePublished - Jun 1 2003
Externally publishedYes

Fingerprint

Kinematics
Monitoring
Sensors
Error detection
Wavelet transforms
Motion analysis

Keywords

  • Ambulatory system
  • Elderly people
  • Kinematic sensor
  • Long-term monitoring
  • Physical activity
  • Postural transition
  • Wavelet transform

ASJC Scopus subject areas

  • Biomedical Engineering

Cite this

Ambulatory system for human motion analysis using a kinematic sensor : Monitoring of daily physical activity in the elderly. / Najafi, Bijan; Aminian, Kamiar; Paraschiv-Ionescu, Anisoara; Loew, François; Büla, Christophe J.; Robert, Philippe.

In: IEEE Transactions on Biomedical Engineering, Vol. 50, No. 6, 01.06.2003, p. 711-723.

Research output: Contribution to journalArticle

Najafi, Bijan ; Aminian, Kamiar ; Paraschiv-Ionescu, Anisoara ; Loew, François ; Büla, Christophe J. ; Robert, Philippe. / Ambulatory system for human motion analysis using a kinematic sensor : Monitoring of daily physical activity in the elderly. In: IEEE Transactions on Biomedical Engineering. 2003 ; Vol. 50, No. 6. pp. 711-723.
@article{15fbbb42897d4889bf72c692d00e21cc,
title = "Ambulatory system for human motion analysis using a kinematic sensor: Monitoring of daily physical activity in the elderly",
abstract = "A new method of physical activity monitoring is presented, which is able to detect body postures (sitting, standing, and lying) and periods of walking in elderly persons using only one kinematic sensor attached to the chest. The wavelet transform, in conjunction with a simple kinematics model, was used to detect different postural transitions (PTs) and walking periods during daily physical activity. To evaluate the system, three studies were performed. The method was first tested on 11 community-dwelling elderly subjects in a gait laboratory where an optical motion system (Vicon) was used as a reference system. In the second study, the system was tested for classifying PTs (i.e., lying-to-sitting, sitting-to-lying, and turning the body in bed) in 24 hospitalized elderly persons. Finally, in a third study monitoring was performed on nine elderly persons for 45-60 min during their daily physical activity. Moreover, the possibility-to-perform long-term monitoring over 12 h has been shown. The first study revealed a close concordance between the ambulatory and reference systems. Overall, subjects performed 349 PTs during this study. Compared with the reference system, the ambulatory system had an overall sensitivity of 99{\%} for detection of the different PTs. Sensitivities and specificities were 93{\%} and 82{\%} in sit-to-stand, and 82{\%} and 94{\%} in stand-to-sit, respectively. In both first and second studies, the ambulatory system also showed a very high accuracy (> 99{\%}) in identifying the 62 transfers or rolling out of bed, as well as 144 different posture changes to the back, ventral, right and left sides. Relatively high sensitivity (> 90{\%}) was obtained for the classification of usual physical activities in the third study in comparison with visual observation. Sensitivities and specificities were, respectively, 90.2{\%} and 93.4{\%} in sitting, 92.2{\%} and 92.1{\%} in {"}standing + walking,{"} and, finally, 98.4{\%} and 99.7{\%} in lying. Overall detection errors (as percent of range) were 3.9{\%} for {"}standing + walking,{"} 4.1{\%} for sitting, and 0.3{\%} for lying. Finally, overall symmetric mean average errors were 12{\%} for {"}standing + walking,{"} 8.2{\%} for sitting, and 1.3{\%} for lying.",
keywords = "Ambulatory system, Elderly people, Kinematic sensor, Long-term monitoring, Physical activity, Postural transition, Wavelet transform",
author = "Bijan Najafi and Kamiar Aminian and Anisoara Paraschiv-Ionescu and Fran{\cc}ois Loew and B{\"u}la, {Christophe J.} and Philippe Robert",
year = "2003",
month = "6",
day = "1",
doi = "10.1109/TBME.2003.812189",
language = "English (US)",
volume = "50",
pages = "711--723",
journal = "IEEE Transactions on Biomedical Engineering",
issn = "0018-9294",
publisher = "IEEE Computer Society",
number = "6",

}

TY - JOUR

T1 - Ambulatory system for human motion analysis using a kinematic sensor

T2 - Monitoring of daily physical activity in the elderly

AU - Najafi, Bijan

AU - Aminian, Kamiar

AU - Paraschiv-Ionescu, Anisoara

AU - Loew, François

AU - Büla, Christophe J.

AU - Robert, Philippe

PY - 2003/6/1

Y1 - 2003/6/1

N2 - A new method of physical activity monitoring is presented, which is able to detect body postures (sitting, standing, and lying) and periods of walking in elderly persons using only one kinematic sensor attached to the chest. The wavelet transform, in conjunction with a simple kinematics model, was used to detect different postural transitions (PTs) and walking periods during daily physical activity. To evaluate the system, three studies were performed. The method was first tested on 11 community-dwelling elderly subjects in a gait laboratory where an optical motion system (Vicon) was used as a reference system. In the second study, the system was tested for classifying PTs (i.e., lying-to-sitting, sitting-to-lying, and turning the body in bed) in 24 hospitalized elderly persons. Finally, in a third study monitoring was performed on nine elderly persons for 45-60 min during their daily physical activity. Moreover, the possibility-to-perform long-term monitoring over 12 h has been shown. The first study revealed a close concordance between the ambulatory and reference systems. Overall, subjects performed 349 PTs during this study. Compared with the reference system, the ambulatory system had an overall sensitivity of 99% for detection of the different PTs. Sensitivities and specificities were 93% and 82% in sit-to-stand, and 82% and 94% in stand-to-sit, respectively. In both first and second studies, the ambulatory system also showed a very high accuracy (> 99%) in identifying the 62 transfers or rolling out of bed, as well as 144 different posture changes to the back, ventral, right and left sides. Relatively high sensitivity (> 90%) was obtained for the classification of usual physical activities in the third study in comparison with visual observation. Sensitivities and specificities were, respectively, 90.2% and 93.4% in sitting, 92.2% and 92.1% in "standing + walking," and, finally, 98.4% and 99.7% in lying. Overall detection errors (as percent of range) were 3.9% for "standing + walking," 4.1% for sitting, and 0.3% for lying. Finally, overall symmetric mean average errors were 12% for "standing + walking," 8.2% for sitting, and 1.3% for lying.

AB - A new method of physical activity monitoring is presented, which is able to detect body postures (sitting, standing, and lying) and periods of walking in elderly persons using only one kinematic sensor attached to the chest. The wavelet transform, in conjunction with a simple kinematics model, was used to detect different postural transitions (PTs) and walking periods during daily physical activity. To evaluate the system, three studies were performed. The method was first tested on 11 community-dwelling elderly subjects in a gait laboratory where an optical motion system (Vicon) was used as a reference system. In the second study, the system was tested for classifying PTs (i.e., lying-to-sitting, sitting-to-lying, and turning the body in bed) in 24 hospitalized elderly persons. Finally, in a third study monitoring was performed on nine elderly persons for 45-60 min during their daily physical activity. Moreover, the possibility-to-perform long-term monitoring over 12 h has been shown. The first study revealed a close concordance between the ambulatory and reference systems. Overall, subjects performed 349 PTs during this study. Compared with the reference system, the ambulatory system had an overall sensitivity of 99% for detection of the different PTs. Sensitivities and specificities were 93% and 82% in sit-to-stand, and 82% and 94% in stand-to-sit, respectively. In both first and second studies, the ambulatory system also showed a very high accuracy (> 99%) in identifying the 62 transfers or rolling out of bed, as well as 144 different posture changes to the back, ventral, right and left sides. Relatively high sensitivity (> 90%) was obtained for the classification of usual physical activities in the third study in comparison with visual observation. Sensitivities and specificities were, respectively, 90.2% and 93.4% in sitting, 92.2% and 92.1% in "standing + walking," and, finally, 98.4% and 99.7% in lying. Overall detection errors (as percent of range) were 3.9% for "standing + walking," 4.1% for sitting, and 0.3% for lying. Finally, overall symmetric mean average errors were 12% for "standing + walking," 8.2% for sitting, and 1.3% for lying.

KW - Ambulatory system

KW - Elderly people

KW - Kinematic sensor

KW - Long-term monitoring

KW - Physical activity

KW - Postural transition

KW - Wavelet transform

UR - http://www.scopus.com/inward/record.url?scp=0037599299&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0037599299&partnerID=8YFLogxK

U2 - 10.1109/TBME.2003.812189

DO - 10.1109/TBME.2003.812189

M3 - Article

VL - 50

SP - 711

EP - 723

JO - IEEE Transactions on Biomedical Engineering

JF - IEEE Transactions on Biomedical Engineering

SN - 0018-9294

IS - 6

ER -