Dear all,

This is my second thread on ‘functional calibration’, this time on knee functional axis (rather than hip functional centre:
I spent quite a bit of time evaluating functional knee axis calibration as a means to define the medio-lateral axis of the femur coordinate system for gait analysis. We have evaluated the different algorithms against freehand 3D ultrasound or EOS bi-plane imaging.

I have also been a long time user of PECS with Vicon Nexus to add kinematics fitting methods to our routine processing for gait analysis in Melbourne. Nexus 2 does not support PECS anymore but provides a new way to interact with Matlab, which I have been playing with recently. So, for those interested in knee functional calibration for gait analysis, you will find here: a Matlab script to run DynaKAD (Baker et al. 1999, called 2DoF) and the Axis Transformation Technique (ATT, Ehrig et al. 2007) and a brief help which explains some pre-requisites for the methods. The methods may be run on any knee flexion-extension movements or directly during walking. At the moment, the 2DoF method does not use any optimised tracking of the segments (e.g. least squares fit of the marker clusters) but the outputs from PiG directly. This is not optimal, but would require a several steps process that may be difficult to share at the moment. I am working on designing a 'user-friendly' alternative.

Hope the script may be useful for you to either understand how to interact with Matlab from Nexus 2 or use, and test, functional knee axis calibration to determine the medio-lateral axis of the femur. Don't hesitate to contact me if you find a bug in the script or you have ideas on how to improve it.
I’ll also be talking at ESMAC 2016 in Seville (Thursday, OS03: Imaging and Anatomy) next week for those interested in the evaluation of these methods.