Development and prevalidation of a measurement tool for the pivot shift phenomenon of the knee

Rupture of the anterior cruciate ligament (ACL) is one of the sports injuries with the highest incidence. The ACL is essential to knee joint stability; its function is to limit anterior subluxation and internal rotation of the tibia relative to the femur. When a rupture occurs, the knee joint presents increased laxity and a rotational instability. During daily activities, this instability often results in a feeling that the knee is slipping, or giving way.;To do so, we first developed a system to attach electromagnetic sensors to the lower limb and record the knee joint kinematics of subjects during the pivot shift test. Two separate data acquisition protocols ensued. In the first, three orthopaedic surgeons each performed the test on twelve different subjects and subjectively graded the pivot shift they produced. The aim of this phase was to develop a method capable of diminishing the variability introduced in the recordings by the clinicians' gestures. In the second phase, an additional 53 subjects were evaluated by one of eight different surgeons. The subjective grades they attributed were used as a gold standard for data analysis. Principal component analysis (PCA) was used to determine which features of the kinematics explain most of the variability between recordings. Using these features, a support vector machine (SYM) based classifier was developed to automatically attribute the pivot shift grade.;The results showed that the variability between clinicians could be diminished by an average of 20% using the velocity of flexion applied by the evaluating clinician. This yielded significant differences between the pivot shift grades for many kinematic parameters. The ACP analysis showed that translation is the most important component of the pivot shift and that its velocity and acceleration are more important than its actual amplitude. Using the most important features, the SYM-based classifier obtained substantial agreement with the clinicians, with 95% of the recordings being classified within one grade of the clinician's attribution.;This thesis has shown that the most important kinematic features of the pivot shift are not those that have traditionally been the focus of quantitative studies. It has also demonstrated the feasibility of objectively attributing the pivot shift grade based on a recording of its kinematics.;Traditional clinical tests measure the increase in knee laxity. Such measures are useful for diagnosing a rupture but they give no indication of the impact of the rupture on knee joint function. The pivot shift test is a clinical test that reproduces and grades the rotational instability following ACL-rupture. It is the only test that has been shown to correlate to subjective criteria of knee joint function such as patient satisfaction, return to physical activity and episodes of giving way. The pivot shift test is graded based on the clinician's interpretation of a subjective scale, rendering it poorly reliable, especially in the hands of a less experienced clinician. There is currently no objective method for grading the pivot shift. The objective of this thesis was therefore to develop such an objective method to attribute the grade based on recorded knee joint kinematics.;Keywords: anterior cruciate ligament, pivot shift, knee joint instability, knee joint kinematics, grade classification...