Motion analysis: Accuracy of joint center estimation (post-doc project Hertha-Firnberg Price T318-N14 of the Austrian Science Fund (FWF)


In the framework of human motion analysis, an exact determination of joint axes and centres is crucial. Errors in the axes’ or centre’s location have an impact on the accuracy of computed joint moments. The kinematic analysis of a movement is distorted by an inappropriate description of a particular joint in the kinematic chain. It is well known from clinical gait analysis, an erroneous description of the knee joint leads to the calculation of unphysiological displacements at the hip joint. This is the result of simultaneous optimization of hip, knee and ankle joint.

With the objective of increasing the reliability of the axis determination, several concurrent procedures have been designed. Two main approaches are distinguished, the first one predicting axes or centres from anatomical landmarks, the second one calculating axes or centres from joint kinematics data. The last named procedures can again be distinguished in general methods applicable to any joint. Among those the concepts of the ‘finite helical axis’, ‘sphere- or cylinder- fitting’ methods or ‘coordinate transformation’ methods are mentioned. In contrast to that, specific methods account for the properties of a joint in particular.

One aim of the present project is to find a parametric model of the human knee joint as the basis for the optimization of joint axes. By accounting for measurement error and noise, it has been argued that a knee with intact ligaments and in the absence of osteoarthritis is best described as a compound hinge joint. The dominant rotation, flexion/extension takes place about the first axis. Simultaneously, a rotation about the tibia’s long axis occurs, the so-called internal-external rotation. This model is presently implemented and validation on real data is intended.


Project leader: Dr. Reichl / contact: