Dynamical analysis of gait motion in osteoarthritic women patients

Abstract

Over 300 million people around the world, mainly women, suffer from Osteoarthritis (OA), a multifactorial disease affecting joints. It occurs most frequently in hands, hips, and knees. The most common solution is having a total knee replacement (TKR) placing a prosthesis on the patient. In the majority of these cases the decision of placing the prosthesis is based not only on objective radiographic measures but also on the pain felt by the patient and its perception, making this decision subjective. Nowadays, studies of human motion dynamics have been frequently applied with biomechanical and computational models that use kinematic and kinetic parameters in order to help clinicians in treatment decision. Some dynamic approaches to gait analysis were also presented, but they were never performed over OA subjects. Nonlinear time series analysis forms a group of algorithms and measures used to extract dynamical features underlying measured signals. It allows to describe dynamical systems where nonlinearities lead to complex time evolution. Unlike deterministic models that produce the same results for a particular set of inputs, stochastic models predict outcomes that account for certain levels of unpredictability or randomness. Using the nonlinear prediction error and a simple irregularity analysis measures we study how predictable the gait will be in the next steps. In this study, human gait recordings of 13 women between 60 to 67 years old that suffer from OA will be analysed from which 6 subjects were referred to take a TKR while the others take a conservative treatment. The aim is to analyse differences in the predictability of the underlying dynamics and its irregularity between TKR and conservative patients. Our results show patients with TKR are more resilient and maintain more coherence compared to conservative patients who seem to present a more stochastic behaviour. Doing so, a quantitative analysis can help clinicians in the treatment decision.