Automatic Non-Invasive Method to Investigate Human Swing of the Gait Cycle

The purpose of this study was to analyse the electromyogram (EMG) signal processing and its application to identify human swing phase of the gait cycle so to develop an automatic, intuitive, low-cost and not invasive EMG device. Electromyography is traditionally used for diagnosis of neuromuscular disorders. EMG signals acquired from muscles require advanced tools and methodologies. However, with the advent of increasingly powerful and low cost sensors and microcontrollers, the study of EMG has found its place in robotics, in the creation of advanced prostheses and in rehabilitation techniques. In this study, a single channel surface EMG signal was studied from human muscles using non-invasive electrodes. After amplification stage, the EMG signal was digitized through analogue and digital (A/D) converter to have accurate and clear signal. Many paediatric neuromuscular disorders are analogous to those seen in the adult and the electrodiagnostic evaluation provides an important extension to the neurological examination. A total of 140 EMG amplitudes were recorded corresponding to 4 lower limb muscles of 35 subjects aged between 12 and 70 years was generated. For each record, the subject made a coordinated paused walk based on a sequence of audible tones to indicate each step. Results show that the amplitudes obtained in each swing sub-phase of the gait records are coherent with the normal swing phase. These findings allow to recommend the use of the EMG acquisition prototype for studies addressed to the detection of motion intention. Future studies will be carried out to extend this methodology to the study of gait in paediatric subject in order to create systems for the diagnosis of deviations from normal gait using non-invasive rehabilitation techniques.