Personalised 3D knee compliance from clinically viable knee laxity measurements: A proof of concept ex vivo experiment.

Personalised information of knee mechanics is increasingly used for guiding knee reconstruction surgery. We explored use of uniaxial knee laxity tests mimicking Lachman and Pivot-shift tests for quantifying 3D knee compliance in healthy and injured knees. Two healthy knee specimens (males, 60 and 88 years of age) were tested. Six-degree-of-freedom tibiofemoral displacements were applied to each specimen at 5 intermediate angles between 0° and 90° knee flexion. The force response was recorded. Six-degree-of-freedom and uniaxial tests were repeated after sequential resection of the anterior cruciate, posterior cruciate and lateral collateral ligament. 3D knee compliance (C6DOF) was calculated using the six-degrees-of-freedom measurements for both the healthy and ligament-deficient knees and validated using a leave-one-out cross-validation. 3D knee compliance (CCT) was also calculated using uniaxial measurements for Lachman and Pivot-shift tests both conjointly and separately. C6DOF and CCT matrices were compared component-by-component and using principal axes decomposition. Bland-Altman plots, median and 40-60th percentile range were used as measurements of bias and dispersion. The error on tibiofemoral displacements predicted using C6DOF was < 9.6% for every loading direction and after release of each ligament. Overall, there was good agreement between C6DOF and CCT components for both the component-by-component and principal component comparison. The dispersion of principal components (compliance coefficients, positions and pitches) based on both uniaxial tests was lower than that based on single uniaxial tests. Uniaxial tests may provide personalised information of 3D knee compliance.