Hello everyone, I am working on a knee joint registration project. According to the experience of predecessors, I have obtained a 4x4 matrix, but I am unable to calculate the kinematics of the knee joint. I am seeking help here.
How can two 4x4 homogeneous transformation matrices (representing the coordinate systems of bone A and bone B, respectively) be converted into the kinematic parameters of bone B relative to bone A? The calculations needed are:
Relative translation (in the coordinate system of bone A)
Do I need to first establish coordinate systems (including the origin) for the two bones, and how would I represent the coordinate systems in the volume?
Yes, establishing coordinate systems is necessary; however, this is partially handled by SlicerAutoscoper. The pre-processing module outputs bone models in the Autoscoper space. The transformation matrices that are output after tracking the bones in SAM are transforms from the Autoscoper space to the global coordinate system as established by the calibration files used in the configuration file. More information about the different coordinate systems (and the transforms that are generated by the pre-processing module to go between spaces) can be found here.
What is not done by SAM is establishing anatomical coordinate systems. We have previously calculated tibiofemoral kinematics from the 4x4 transformation matrices by resolving six-degree-of-freedom Euler angles and expressing the 3D motion using anatomical coordinate systems. Our methods are described here.
I can understand what you said about the pre-processing module outputting the skeletal model. After looking at the methods you mentioned, I am still not very clear on how to connect the anatomical coordinate system with the 4x4 matrix and obtain tibiofemoral joint kinematics. I would be very grateful if you could provide a more detailed explanation.
