Classification of Wrist Movements from Electromyography Signals

Carpal Tunnel Syndrome (CTS) affects about 5% of the population worldwide. One of the biggest root causes is associated with occupational hazard. Repeated incorrect movement of the wrist joint leads to chronic inflammation which in turn results in CTS. Correct intervention provided at the right time may help in preventing progression or worsening of this condition. While corrective intervention such as physiotherapy is important when symptoms are noticeable, preventive intervention can be considered to avoid an onset of the condition, especially in vulnerable occupations such as those requiring the use of a computer keyboard. This research aims to propose a method of such a preventive intervention. Electromyography (EMG) is a procedure that evaluates the condition of muscles and the nerve cells (motor neurons) that control them. Neurons transmit electrical signals that cause muscles to contract and relax. These signals are recorded by the surface EMG (sEMG) and visualized using a signal processor. This work evaluates the use of decision trees to analyze sEMG signals to recognize specific wrist movements. This approach may have applications in wearables to provide a warning when a wrong movement is made, thus helping prevent incidence and progression of CTS.