Accuracy of muscle moment arms estimated from MRI-based musculoskeletal models of the lower extremity

Abstract

Objective: Biomechanical models that compute the lengths and moment arms of soft tissues are broadly applicable to the treatment of movement abnormalities and the planning of orthopaedic surgical procedures. The goals of this study were to: (i) develop methods to construct subject-specific biomechanical models from magnetic resonance (MR) images, (ii) create models of three lower-extremity cadaveric specimens, and (iii) quantify the accuracy of muscle-tendon lengths and moment arms estimated using these models.

Materials and methods: Models describing the paths of the medial hamstrings and psoas muscles for a wide range of body positions were developed from MR images in one joint configuration by defining kinematic models of the hip and knee, and by specifying "wrapping surfaces" that simulate interactions between the muscles and underlying structures. Our methods for constructing these models were evaluated by comparing hip and knee flexion moment arms estimated from models of three specimens to the moment arms determined experimentally on the same specimens. Because a muscle's moment arm determines its change in length with joint rotation, these comparisons also tested the accuracy with which the models could estimate muscle-tendon lengths over a range of hip and knee motions.

Results: Errors in the moment arms calculated with the models, averaged over functional ranges of hip and knee flexion, were less than 4 mm (within 10% of experimental values).

Conclusion: The combination of MR imaging and graphics-based musculoskeletal modeling provides an accurate and efficient means of estimating muscle-tendon lengths and moment arms in vivo.