Determination of gait events using an externally mounted shank accelerometer

Jonathan Sinclair, Sarah J Hobbs, Laurence Protheroe, Christopher J Edmundson, Andrew Greenhalgh

Research output: Contribution to journalArticlepeer-review

43 Citations (Scopus)


Biomechanical analysis requires the determination of specific foot contact events. This is typically achieved using force platform information; however, when force platforms are unavailable, alternative methods are necessary. A method was developed for the determination of gait events using an accelerometer mounted to the distal tibia, measuring axial accelerations. The aim of the investigation was to determine the efficacy of this method. Sixteen participants ran at 4.0 m/s ± 5%. Synchronized tibial accelerations and vertical ground reaction forces were sampled at 1000 Hz as participants struck a force platform with their dominant foot. Events determined using the accelerometer, were compared with the corresponding events determined using the force platform. Mean errors of 1.68 and 5.46 ms for average and absolute errors were observed for heel strike and of -3.59 and 5.00 ms for toe-off. Mean and absolute errors of 5.18 and 11.47 ms were also found for the duration of the stance phase. Strong correlations (r = .96) were also observed between duration of stance obtained using the two different methods. The error values compare favorably to other alternative methods of predicting gait events. This suggests that shank-mounted accelerometers can be used to accurately and reliably detect gait events.

Original languageEnglish
Pages (from-to)118-22
Number of pages5
JournalJournal of Applied Biomechanics
Issue number1
Publication statusPublished - 1 Feb 2013


  • Acceleration
  • Adult
  • Equipment Design
  • Equipment Failure Analysis
  • Female
  • Foot
  • Gait
  • Humans
  • Locomotion
  • Male
  • Monitoring, Ambulatory
  • Reproducibility of Results
  • Sensitivity and Specificity
  • Journal Article


Dive into the research topics of 'Determination of gait events using an externally mounted shank accelerometer'. Together they form a unique fingerprint.

Cite this