Childhood conditions such as developmental hip abnormalities, cerebral palsy and slipped capital femoral epiphysis can lead to complex hip and knee deformities. Due to the variability in paediatric conditions, outcomes of surgical interventions are variable making it difficult to quantify efficacy of individual treatments, or know which are most effective for which child. Computational modelling plays an important role in our understanding of musculoskeletal form and function, however modelling has yet to play a role in routine clinical care. Models’ ability to predict clinical outcomes is dependent on accurately capturing patients’ complex geometry of the musculoskeletal system and describing muscle and joint forces. Image-based subject-specific models of the neuro-musculoskeletal system are capable of estimating joint loads and show promises for predicting clinical outcomes. However, these models are time-consuming, costly, and requires high levels of expertise. Solving this challenge is a significant step toward bringing computational modelling into the clinic.