A commentary on Kalkman et al.’s letter to the editor regarding Alexander et al. (2019): “Children with cerebral palsy have larger in-vivo and linearly scaled Achilles tendon moment arms than typically developing children”

Rather than focusing this communication on comparing and contrasting specific modelling differences between an established 2D AT method (Kalkman et al. 2017) and a valid 3D in-vivo AT MA method (Alexander et al., 2017), our goal is to highlight the global differences between these methods, which we hope will help guide and progress future researcher within this important research area (i.e., paediatric CP and ankle joint & muscuolotendon modelling). The accurate estimates of muscle forces are reliant on accurate AT MA estimates, which are non-linearly influenced by four interconnected musculoskeletal modelling parameters: (1) musculotendon properties (e.g., fibre length, tendon slack length, etc.), (2) musculotendon geometry (e.g., insertion points, volume, etc.), (3) bony morphology (e.g., shape, deformities etc.) and (4) joint geometry (i.e., axes, degrees of freedom, condylar surface, etc.). Though time, cost and computational restrictions have limited the integration of medical imaging within standard motion capture and musculoskeletal modelling frameworks, it is the next frontier for the field of musculoskeletal modelling to overcome if it is to continue to add value to clinical best practice research and decision making.