Postural control modeling: bridging neural activity and COP oscillations

Postural control relies on the integration of sensory inputs from the visual, vestibular and proprioceptive systems. While traditional studies focus on biomechanical measures, such as the Centre of Pressure (COP) displacement, the neural mechanisms underlying balance regulation remain scarcely explored. This dissertation consists of developing a neural network model to simulate cerebellar and motor cortex activity under different Sensory Organization Test (SOT) conditions, analysing neural activation, related barycentre shifts and Sample Entropy (SampEn). Results show that as task difficulty increases, cerebellar SampEn decreases, suggesting a more constrained selection of motor plans, while motor SampEn follows a trend similar to the COP’s entropy found in scientific literature, reflecting increased irregularity in postural adjustments. These findings highlight the potential of non-linear analyses in evaluating neural activity and suggest a close link between neural and biomechanical data. This work demonstrated the feasibility of neural network models for investigating postural control mechanisms. Future research should validate these models with experimental neural recordings, such as EEG, to further refine our understanding of balance regulation.