Parameter Uncertainty in Multibody Models of the Natural Knee Joint: A Scoping Review

Background: Multibody models are essential for studying knee joint mechanics, but their reliability and subsequent clinical utility are limited by uncertainties in ligament and contact parameters. Currently, no consensus exists on which parameters to prioritize or which statistical distributions best establish model credibility. Objectives: This scoping review aims to systematize reported uncertainty values for ligament and contact parameters in multibody models of the natural knee to identify trends and research gaps. Methods: Following PRISMA-ScR guidelines, a systematic search was conducted across PubMed, Scopus, and Web of Science. Methodological quality was assessed using a customized 13-item checklist, and the data were synthesized via a narrative approach by charting parameter types, quantification methods, and model structures. Results: In total, 19 articles were included (out of 494 identified), showing a wide variability in uncertain parameter types, values, and modeling approaches. Ligaments were typically represented as deformable cables with quadratic–linear behavior, while articular contact utilized elastic foundation formulations or mechanisms. Standard deviations of 30% of the mean for ligament stiffness and 0.02 for reference strain (typically modeled within Gaussian distributions) were the most frequently quantified uncertain parameters. Geometric uncertainties for ligament attachment points varied widely, ranging from 1.0 to 5.0 mm. Idealized contact geometry also varied within 2.5 mm for linear coordinates and 15° for angular coordinates. Conclusions: Wide variability and inconsistent reports highlight a need for standardized definitions of parameter uncertainty in multibody knee modeling to improve reproducibility of musculoskeletal knee simulations and ensure a reliable transition of these models into clinical practice.