Magnetic resonance imaging of patellofemoral relationships

Abstract

Patellofemoral relationships were analyzed in 11 patients (13 knees) with patellar dislocation and 15 asymptomatic subjects (15 knees) at 0° and 20° of flexion. The measurements were made from five consecutive axial images through the patellofemoral joint. The six indices measured were lateral patellar tilt (LPT), lateral patellofemoral angle (LPA), lateral patellar displacement (LPD), patella-lateral condyle index (L/PW), congruence angle (CA), and sulcus angle (SA). The reproducibility of the method was evaluated. The difference between the two study groups was more evident at 0° than at 20° of knee flexion. Significant differences were noted between measurements made at different levels of the joint, particularly in the controls. Isometric contraction of the quadriceps muscle lateralized and tilted the patella slightly in both groups. L/PW with and without quadriceps muscle contraction, and LPA with reference to the anterior condyles differentiated between the two study groups most clearly. LPT and LPA with reference to the anterior condyles differentiated the study groups better than LPT and LPA with reference to the posterior condyles. The reproducibility was good except for inter-observer comparison of CA and SA. The use of an imaging plane selected at the midpoint of the patellar articular cartilage increases the sensitivity of the measurements, since it takes into account both the height of the patella and the tendency towards lateralization. These results indicate that patellar tilt is best measured with the LPA index and patellar lateralization with the L/PW index at 0° knee flexion. This study should always include isometric contraction of the quadriceps muscle.