ABSTRACT
By restricting analysis to single averaged strides considered to be characteristic for the individual under investigation, current methods in gait analysis do not exploit the full dynamics of continuous locomotion. Therefore, a novel approach is presented that is based on long-term measurements of kinematic data during treadmill walking. The method consists of reconstructing the system attractor in the embedding space and then analyzing its geometric structure. Estimating the dimension of movement trajectories correlates well with the notion of controlling multiple degrees of freedom during performance of complex movement tasks such as walking. The influence of walking speed on the complexity of physiologic walking was investigated in 10 healthy subjects walking on a treadmill at seven fixed speeds. The results suggest that human walking becomes more complex at slower speeds. This may be associated with results from EMG studies demonstrating more irregular EMG patterns at very slow walking speeds. This study emphasizes that tools from non-linear dynamics are well suited for providing more insight into motor control in humans.
Subject(s)
Biomechanical Phenomena/methods , Joints/physiology , Leg/physiology , Models, Biological , Pattern Recognition, Automated/methods , Range of Motion, Articular/physiology , Walking/physiology , Algorithms , Computational Biology/methods , Computer Simulation , Diagnosis, Computer-Assisted/methods , Gait/physiology , Humans , Signal Processing, Computer-AssistedABSTRACT
Recently, locomotion therapy on a treadmill has become part of rehabilitation programs for neurological gait disorders (spinal cord injury, hemiplegia). Instrumental gait analysis is an important tool for quantification of therapy progress in terms of functional changes in a patient's gait patterns. Whereas most of current applications focus on diagnostic assessment of gait, the current paper presents an extension of movement analysis offering enhanced therapeutic options. Specifically, the conceptual basis for application of real-time movement feedback in rehabilitation medicine is outlined and is put into context with recent developments in the field. A first technical realization of these concepts is presented and first results are reported. Furthermore, open questions towards a universal environment for movement feedback in rehabilitation medicine are discussed and future lines of research are identified.