Understanding Foot Function During Stance Phase by Bayesian Network Based Causal Inference

Abstract

Listen

Translate article

Understanding the biomechanics of the human foot during each stage of walking is important for the objective evaluation of movement dysfunction, accuracy of diagnosis, and prediction of foot impairment. Extracting causal relations from amongst the muscle activities, toe trajectories, and plantar pressures during walking assists in recognizing several disease conditions, and understanding the hidden complexity of human foot functions, thus, facilitating appropriate therapy and treatment. To extract these relations, we applied the Bayesian Network (BN) model to data collected in the stance phase of walking. For a better understanding of foot function, the experimental data were divided into three stages (initial contact, loading response to mid-stance, and terminal stance to pre-swing). BNs were constructed for these three stages of data for normal walking and simulated hemiplegic walking, then compared and analyzed. Results showed that BNs extracted could express the underlying mechanism of foot function.KeywordsBayesian networkStance phaseMuscle activitiesFoot function