Motion-based tendon diagnosis using sequence processing of ultrasound images.

ABSTRACT

Degenerative or pre-degenerative processes in the tendon, which could lead to a spontaneous rupture, are well known problems, especially among athletes. Existing non-invasive diagnostic methods are mainly based on subjective analysis of static images of the tendon, but in many cases the degeneration cannot be diagnosed in time. Combining a set of existing image processing techniques, a tool for tracking the in vivo motion of a tendon imaged with dynamic ultrasound was implemented. A group of subjects that had undergone a degenerative rupture of their Achilles tendon, one subject with a traumatic rupture, and a group of control subjects were all tested. Using the motion information that was obtained from both tendons of all subjects, we developed an automatic test that examines the symmetrical properties of the tendon's motion, and defined a negative asymmetry property that could be quantified as a score. This score was found to be significantly more enhanced in the post-operative tendons (18.0 +/- 9.0) than in the contra-lateral healthy tendons of the same subjects (3.9 +/- 4.6). In the single traumatic rupture subject, this effect was not found (0.0). This leads us to believe that the negative asymmetry of tendon motion may be associated with degenerative or pre-degenerative processes in the tendon. Also, the mean degree of negative asymmetry in the healthy tendons of post-operative subjects (3.9 +/- 4.6) was found to be higher than that of healthy tendons of the control group (1.5 +/- 1.8). This finding may be associated with the fact that tendons that are contra-lateral to spontaneously ruptured tendons have a higher risk of developing degenerative processes. The method presented here is objective, low-cost, non-invasive and possibly more sensitive than existing non-invasive techniques.