Measurement of the screw-home motion of the knee is sensitive to errors inaxis alignment

Measurements of joint angles during motion analysis are subject to error ca used by kinematic crosstalk, that is, one joint rotation (e.g., flexion) be ing interpreted as another (e.g., abduction). Kinematic crosstalk results f rom the chosen joint coordinate system being misaligned with the axes about which rotations are assumed to occur. The aim of this paper is to demonstr ate that measurement of the so- called "screw-home" motion of the human kne e, in which axial rotation and extension are coupled, is especially prone t o errors due to crosstalk. The motions of two different two-segment mechani cal linkages were examined to study the effects of crosstalk. The segments of the first linkage (NSH) were connected by a revolute joint, but the seco nd linkage (SH) incorporated gearing that caused 15 degrees of screw-home r otation to occur with 90 degrees knee flexion. It was found that rotating t he flexion axis (inducing crosstalk) could make linkage NSH appear to exhib it a screw-home motion and that a different rotation of the flexion axis co uld make linkage SH apparently exhibit pure flexion. These findings suggest that the measurement of screw-home rotation may be strongly influenced by errors in the location of the flexion axis. The magnitudes of these displac ements of the flexion axis were consistent with the inter-observer variabil ity seen when five experienced observers defined the flexion axis by palpat ing the medial and lateral femoral epicondyles. Care should be taken when i nterpreting small internal-external rotations and abduction-adduction angle s to ensure that they are not the products of kinematic crosstalk. (C) 2000 Elsevier Science Ltd. All rights reserved.