Details

Autor(en) / Beteiligte

• Binnie, Tara
• Smith, Anne
• Kent, Peter
• Ng, Leo
• O’Sullivan, Peter
• Tan, Jay-Shian
• Davey, Paul C.
• Campbell, Amity

Titel

Concurrent validation of inertial sensors for measurement of knee kinematics in individuals with knee osteoarthritis: A technical report

Ist Teil von

• Health and technology, 2022, Vol.12 (1), p.107-116

Ort / Verlag

Berlin/Heidelberg: Springer Berlin Heidelberg

Erscheinungsjahr

2022

Quelle

Alma/SFX Local Collection

Beschreibungen/Notizen

• The validity of inertial sensor-based kinematic measurements in individuals with knee osteoarthritis has not previously been investigated. This study assessed the concurrent validity of inertial sensors in measuring knee kinematics in individuals with knee osteoarthritis when compared to the Vicon motion analysis system, and explored the difference between two calculation methods (traditional approach versus functional approach). Nineteen participants with knee osteoarthritis performed functional tasks with DorsaVi sensors (fixed with Vicon markers) worn on thighs and shanks. Peak and time-series knee flexion and extension (for both DorsaVi and Vicon systems) were calculated using the two calculation methods. Agreement between the systems was estimated by calculating root mean squared errors, mean differences and 95% limits of agreement. For the traditional approach, the root mean squared error between the DorsaVi and Vicon measurements ranged from 1.70°-3.02° for peak and 3.72–4.67° for time-series knee flexion and extension. For the functional approach, the root mean squared error ranged from 1.77°-3.18° for peak and 4.58°-5.04° for time-series knee flexion and extension. The mean difference varied across tasks (traditional approach: -0.16°-3.76°, functional approach: 0.96°-4.94°), and the limits of agreement showed high variability between the DorsaVi and Vicon measurements across the sample. Although there appears to be acceptable agreement between the systems for measuring knee kinematics, there was high variability in measurement differences across the dataset. In addition, a functional calibration approach does not appear to improve the accuracy of inertial sensor-based knee kinematics.