Microsoft Kinect can distinguish differences in over-ground gait between older persons with and without Parkinson's disease

Moataz Mohamed Eltoukhy, Christopher Kuenze, Jeonghoon Oh, Marco Jacopetti, Savannah Wooten, Joseph Signorile

Research output: Contribution to journalArticle

21 Citations (Scopus)

Abstract

Gait patterns differ between healthy elders and those with Parkinson's disease (PD). A simple, low-cost clinical tool that can evaluate kinematic differences between these populations would be invaluable diagnostically; since gait analysis in a clinical setting is impractical due to cost and technical expertise. This study investigated the between group differences between the Kinect and a 3D movement analysis system (BTS) and reported validity and reliability of the Kinect v2 sensor for gait analysis. Nineteen subjects participated, eleven without (C) and eight with PD (PD). Outcome measures included spatiotemporal parameters and kinematics. Ankle range of motion for C was significantly less during ankle swing compared to PD (p = 0.04) for the Kinect. Both systems showed significant differences for stride length (BTS (C 1.24 ± 0.16, PD= 1.01 ± 0.17, p = 0.009), Kinect (C= 1.24 ± 0.17, PD= 1.00 ± 0.18, p = 0.009)), gait velocity (BTS (C= 1.06 ± 0.14, PD= 0.83 ± 0.15, p = 0.01), Kinect (C= 1.06 ± 0.15, PD= 0.83 ± 0.16, p = 0.01)), and swing velocity (BTS (C= 2.50 ± 0.27, PD= 2.12 ± 0.36, p = 0.02), Kinect (C= 2.32 ± 0.25, PD= 1.95 ± 0.31, p = 0.01)) between groups. Agreement (RangeICC = 0.93–0.99) and consistency (RangeICC = 0.94–0.99) were excellent between systems for stride length, stance duration, swing duration, gait velocity, and swing velocity. The Kinect v2 can was sensitive enough to detect between group differences and consistently produced results similar to the BTS system.

Original languageEnglish (US)
Pages (from-to)1-7
Number of pages7
JournalMedical Engineering and Physics
Volume44
DOIs
StatePublished - Jun 1 2017

Fingerprint

Gait
Parkinson Disease
Gait analysis
Biomechanical Phenomena
Ankle
Kinematics
Professional Competence
Costs and Cost Analysis
Articular Range of Motion
Reproducibility of Results
Costs
Outcome Assessment (Health Care)
Sensors

Keywords

  • Gait analysis
  • Kinect
  • Motion capture
  • Parkinson's disease

ASJC Scopus subject areas

  • Biophysics
  • Biomedical Engineering

Cite this

Microsoft Kinect can distinguish differences in over-ground gait between older persons with and without Parkinson's disease. / Eltoukhy, Moataz Mohamed; Kuenze, Christopher; Oh, Jeonghoon; Jacopetti, Marco; Wooten, Savannah; Signorile, Joseph.

In: Medical Engineering and Physics, Vol. 44, 01.06.2017, p. 1-7.

Research output: Contribution to journalArticle

@article{8e4fb1afb3894bfdada5f1f2caa40600,
title = "Microsoft Kinect can distinguish differences in over-ground gait between older persons with and without Parkinson's disease",
abstract = "Gait patterns differ between healthy elders and those with Parkinson's disease (PD). A simple, low-cost clinical tool that can evaluate kinematic differences between these populations would be invaluable diagnostically; since gait analysis in a clinical setting is impractical due to cost and technical expertise. This study investigated the between group differences between the Kinect and a 3D movement analysis system (BTS) and reported validity and reliability of the Kinect v2 sensor for gait analysis. Nineteen subjects participated, eleven without (C) and eight with PD (PD). Outcome measures included spatiotemporal parameters and kinematics. Ankle range of motion for C was significantly less during ankle swing compared to PD (p = 0.04) for the Kinect. Both systems showed significant differences for stride length (BTS (C 1.24 ± 0.16, PD= 1.01 ± 0.17, p = 0.009), Kinect (C= 1.24 ± 0.17, PD= 1.00 ± 0.18, p = 0.009)), gait velocity (BTS (C= 1.06 ± 0.14, PD= 0.83 ± 0.15, p = 0.01), Kinect (C= 1.06 ± 0.15, PD= 0.83 ± 0.16, p = 0.01)), and swing velocity (BTS (C= 2.50 ± 0.27, PD= 2.12 ± 0.36, p = 0.02), Kinect (C= 2.32 ± 0.25, PD= 1.95 ± 0.31, p = 0.01)) between groups. Agreement (RangeICC = 0.93–0.99) and consistency (RangeICC = 0.94–0.99) were excellent between systems for stride length, stance duration, swing duration, gait velocity, and swing velocity. The Kinect v2 can was sensitive enough to detect between group differences and consistently produced results similar to the BTS system.",
keywords = "Gait analysis, Kinect, Motion capture, Parkinson's disease",
author = "Eltoukhy, {Moataz Mohamed} and Christopher Kuenze and Jeonghoon Oh and Marco Jacopetti and Savannah Wooten and Joseph Signorile",
year = "2017",
month = "6",
day = "1",
doi = "10.1016/j.medengphy.2017.03.007",
language = "English (US)",
volume = "44",
pages = "1--7",
journal = "Medical Engineering and Physics",
issn = "1350-4533",
publisher = "Elsevier BV",

}

TY - JOUR

T1 - Microsoft Kinect can distinguish differences in over-ground gait between older persons with and without Parkinson's disease

AU - Eltoukhy, Moataz Mohamed

AU - Kuenze, Christopher

AU - Oh, Jeonghoon

AU - Jacopetti, Marco

AU - Wooten, Savannah

AU - Signorile, Joseph

PY - 2017/6/1

Y1 - 2017/6/1

N2 - Gait patterns differ between healthy elders and those with Parkinson's disease (PD). A simple, low-cost clinical tool that can evaluate kinematic differences between these populations would be invaluable diagnostically; since gait analysis in a clinical setting is impractical due to cost and technical expertise. This study investigated the between group differences between the Kinect and a 3D movement analysis system (BTS) and reported validity and reliability of the Kinect v2 sensor for gait analysis. Nineteen subjects participated, eleven without (C) and eight with PD (PD). Outcome measures included spatiotemporal parameters and kinematics. Ankle range of motion for C was significantly less during ankle swing compared to PD (p = 0.04) for the Kinect. Both systems showed significant differences for stride length (BTS (C 1.24 ± 0.16, PD= 1.01 ± 0.17, p = 0.009), Kinect (C= 1.24 ± 0.17, PD= 1.00 ± 0.18, p = 0.009)), gait velocity (BTS (C= 1.06 ± 0.14, PD= 0.83 ± 0.15, p = 0.01), Kinect (C= 1.06 ± 0.15, PD= 0.83 ± 0.16, p = 0.01)), and swing velocity (BTS (C= 2.50 ± 0.27, PD= 2.12 ± 0.36, p = 0.02), Kinect (C= 2.32 ± 0.25, PD= 1.95 ± 0.31, p = 0.01)) between groups. Agreement (RangeICC = 0.93–0.99) and consistency (RangeICC = 0.94–0.99) were excellent between systems for stride length, stance duration, swing duration, gait velocity, and swing velocity. The Kinect v2 can was sensitive enough to detect between group differences and consistently produced results similar to the BTS system.

AB - Gait patterns differ between healthy elders and those with Parkinson's disease (PD). A simple, low-cost clinical tool that can evaluate kinematic differences between these populations would be invaluable diagnostically; since gait analysis in a clinical setting is impractical due to cost and technical expertise. This study investigated the between group differences between the Kinect and a 3D movement analysis system (BTS) and reported validity and reliability of the Kinect v2 sensor for gait analysis. Nineteen subjects participated, eleven without (C) and eight with PD (PD). Outcome measures included spatiotemporal parameters and kinematics. Ankle range of motion for C was significantly less during ankle swing compared to PD (p = 0.04) for the Kinect. Both systems showed significant differences for stride length (BTS (C 1.24 ± 0.16, PD= 1.01 ± 0.17, p = 0.009), Kinect (C= 1.24 ± 0.17, PD= 1.00 ± 0.18, p = 0.009)), gait velocity (BTS (C= 1.06 ± 0.14, PD= 0.83 ± 0.15, p = 0.01), Kinect (C= 1.06 ± 0.15, PD= 0.83 ± 0.16, p = 0.01)), and swing velocity (BTS (C= 2.50 ± 0.27, PD= 2.12 ± 0.36, p = 0.02), Kinect (C= 2.32 ± 0.25, PD= 1.95 ± 0.31, p = 0.01)) between groups. Agreement (RangeICC = 0.93–0.99) and consistency (RangeICC = 0.94–0.99) were excellent between systems for stride length, stance duration, swing duration, gait velocity, and swing velocity. The Kinect v2 can was sensitive enough to detect between group differences and consistently produced results similar to the BTS system.

KW - Gait analysis

KW - Kinect

KW - Motion capture

KW - Parkinson's disease

UR - http://www.scopus.com/inward/record.url?scp=85017381449&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85017381449&partnerID=8YFLogxK

U2 - 10.1016/j.medengphy.2017.03.007

DO - 10.1016/j.medengphy.2017.03.007

M3 - Article

VL - 44

SP - 1

EP - 7

JO - Medical Engineering and Physics

JF - Medical Engineering and Physics

SN - 1350-4533

ER -