The potential of machine learning in classifying relapse and non-relapse in children with clubfoot based on movement patterns.

The diverse nature and timing of a clubfoot relapse pose challenges for early detection. A relapsed clubfoot typically involves a combination of deformities affecting a child's movement pattern across multiple joint levels, formed by a complex kinematic chain. Machine learning algorithms have the capacity to analyse such complex nonlinear relationships, offering the potential to train a model that assesses whether a child has relapsed clubfoot based on their movement pattern. Hence, this study aimed to explore to what extent biomechanical data collected with three-dimensional movement analysis can be used to classify children with relapsed clubfoot from children with non-relapsed clubfoot. The findings demonstrated the potential of subject classification based on kinematic movement patterns, where combining dynamic activities improves sensitivity in distinguishing children with relapsed clubfoot from children with non-relapsed clubfoot. Moreover, the study highlights biomechanical features that should be considered during clinical follow-up of children with clubfoot. This might aid early identification and treatment of relapsed clubfoot, which is expected to prevent the necessity of surgical treatment in these young patients. However, for future application of machine learning classification in clinical practice, a larger subject population will be necessary to develop a generalizable and robust model.