Development of a Method to Correlate Anatomic Landmarks with CT Data for Biomechanical in Vitro Experiments

A majority of the biomechanical in-vitro studies depend on virtual coordinate systems. These are in most cases based on anatomic landmarks. Typically, the vertebral body is marked by passive or active markers and tracked with a motion capture. Each vertebra needs to be referenced with a digitizing probe. The precision of this registration and the consecutive orientation of the body coordinate system are mainly depended on distinguishable anatomic landmarks. Hence, referencing of vertebral bodies is quite error prone. Further problems arise if the reference procedure has to be redone during testing process or the anatomic landmarks are not accessible because of the operative procedure or testing set-up.Material and Methods To address the problem of repeatability, we developed a set of small marker screws that have specially designed screw heads ( Fig. 1 ). The spherical tip of the digitizing probe can be fitted into the cutout cone of the screw head with the result that the referenced point is equivalent to the point of intersection of the screw axis and the perpendicular base of the cone. Accordingly, if the screws are firmly attached to the bone, it will be possible to reference the points very reliable. The following steps did this: 1st step: Insertion of three or four marker screws into each vertebral body. The positions can be freely chosen as long as the motion capture system is able to generate a coordinate system. Any interference with the operative procedure or testing setup should be prevented at this point. 2nd step: Performing a CT/MRI scan of the specimen. In most cases, this has to be done to exclude compromising defects anyway and consequently, no extra effort is necessary. 3rd step: Referencing of the body coordinate systems with the aid of the marker screws. This can be repeated during the testing procedure if necessary. 4th step: Completion of the in vitro testing procedure.Results Transformation matrices could correct the virtual shift from the referenced body coordinate systems to the anatomically aligned coordinate systems taken from CT scan. This was done by determination of the marker screw heads in the depicted volume and virtual regeneration of the referenced body coordinate systems from the 3rd step. Then the coordinates for the generation of the anatomically aligned coordinate systems can be obtained. On the basis of the two coordinate systems, a transformation matrix can be calculated to shift data from one system to the other. In consequence, it was possible to apply each transformation matrix to the data set obtained during the in vitro testing procedure to shift it virtually to the respective anatomically aligned coordinate system.Conclusion The presented algorithm facilitates the link between the kinetic data from in vitro testing procedures and the precise anatomical data from 3D imaging. Accordingly, it might help to understand complex motion parameters (e.g., FHA) by plotting them into the 3D volume. [Figure: see text]