Methods

We propose to compare these reconstructions thanks to an index of articular coherence. This index measures the instantaneous state surface configurations during a motion. More specifically, this refers to the position between two articular surfaces facing each other. First of all, the index has to refer to a position known to be physiological. This initial position of the bones, named reference, directly results from the segmentation of CT scans. First we compute all distances between the two surfaces and then we compute the Cumulative Distribution Function (CDF). We process this way for each iteration of the motion. Then we obtain a batch of CDF curves which provide us qualitative information relative to the motion such as potential collisions or dislocations. The graph of all CDF curves is called Figure of Articular Coherence (FoAC). A good articular coherence is characterised by CDF which are close to the reference. This qualitative method is coupled to a quantitative one named Index of Articular Coherence (IoAC) which computes the Haussdorff distance between the temporal distributions and the reference. This distance has to be as low as possible. The tools were tested on cadaveric experiments of ACL reconstruction provided by Hagemeister et al, (1999). They recorded the knee flexion/extension motion in following situations: the intact knee, after ACL resection, after three methods of ACL reconstruction on the same knee (‘over-the-top’ method (OTT), two different two tunnel reconstructions (2 tunnel). Our method was used, for the time being, for one specimen. We compare different post-surgery kinematics thanks to the FoAC and IoAC.