Reliability of joint kinematic calculations based on direct kinematic and inverse kinematic models in obese children

Abstract

Reliability values showed clinically acceptable error-margins and were comparable between all models. Therefore, our results support the careful use of IK models in overweight or obese populations, e.g. for musculoskeletal modelling studies. The inconsistent kinematic output can mainly be explained by different model conventions and anatomical segment coordinate frame definitions.

Effectiveness of Global Optimisation and Direct Kinematics in Predicting Surgical Outcome in Children with Cerebral Palsy

“…In the PiG, the pelvis anatomical coordinate system is oriented such that the anteroposterior axis lies in a plane defined by the pelvis markers (sacrum and two anterior iliac spines), whereas the axis for the mGen is oriented parallel to the horizontal. This led to a RMSD difference of about 10-15 • , which was shown to reduce by adjustment of the anatomical and joint reference frames [10,11]. Similarly, in this study, an RMSD of 18 ± 1.5 • in pelvic tilt between the PiG and mGen approaches was found.…”

“…This is most likely to be attributed to the differences in degrees of freedom observed between the two models. In fact, Horsak et al [11] showed in a group of obese children that using three instead of one degree of freedom at the knee joint in mGen reduced the RMSD between mGen and PiG estimates of hip rotation. From their results, it could also be observed that the R 2 improved with a 3-DoF mGen knee.…”

“…It is also recognised that using the consensus clinical judgement with no gait analysis metrics input as a proxy gold standard is by nature subjective, and this may have introduced additional error. In addition, the choice of using the same MCID in GPS of 1.6 • for the mGen, could be considered arbitrary; however, this was justified as akin to using the 5 • threshold for detecting clinically relevant changes in clinical gait analysis kinematics in this and other studies [10,11,14,39]. Finally, future studies should include a larger cohort to verify the generalisability of the reported results.…”

“…This method simultaneously determines the orientation of a constrained skeletal model while minimising measurement errors in marker locations. It addresses some of the challenges of the PiG model, in particular minimising skin tissue artefacts [9], and has been shown to provide reliable estimates of gait kinematics [10,11]. However, its successful adoption in routine assessment of pathological gait is yet to be determined due to limited implementation in commercial software packages and limitations associated to the simplifications made within the reference joint models, such as the use of minimal degrees of freedom as well as complexity and time constraints [12,13].…”

Inter-session repeatability of markerless motion capture gait kinematics

“…We believe instead that the greater repeatability of the markerless kinematics compared to those from previous studies is more likely due to the markerless tracking method itself. We acknowledge that greater differences may be seen if the knee and hip joints were additionally allowed 6DOF; however, other studies have shown similar reliability between IK and DK techniques and there is currently no implemented option for 6DOF knee and hip joints in Theia3D (Mentiplay and Clark, 2018;Horsak et al, 2018).…”