Research On Wearable "EMG-Bridge" System For Upper Extremity Rehabilitation

The limb paralysis caused by stroke and spinal cord injury is a serious disability disease.Such disease affects physical and mental health of human,which demands the research on the rehabilitation and treatment of limb paralysis.Our group has proposed a method called "Electromyographic Bridge"(EMGB)based on functional electrical stimulation(FES).EMGB detects and processes the surface electromyography(s EMG)from the healthy limb.The pattern recognition of EMGB generates the FES pulses to rehabilitate the limb motor function.The advanced wearable medical devices provide convenience for doctors and patients.In this thesis,we proposed a wearable solution for EMGB system.It achieves low power consumption,small form factor.The district contributions to the state-of-the-art within this thesis and summaries as follows,1.Design of the architecture of wearable EMGB system: The architecture consists of 3 main modules.The first module is a wearable 8-channel armband designed to detect s EMG signal.The second module is a real-time s EMG recognition module to recognize the pattern of four wrist movements.The final module is a wearable 4-channel FES module which was designed to generate FES pulses according to the parameters.The limb was activated by FES pluses performing four actions: grasping,wrist extension,wrist flexion,and five-finger extension.2.Design of wearable s EMG detecting armband: Wearable s EMG armband with low power and small size was designed to detect s EMG from skin.It provides 8 detecting channels,each channel has a s EMG sensors and the channels are connected by cables.3.Design of real-time s EMG recognition module: Five time domain features of 8-channel surface EMG and LDA were used to perform real-time recognition of four actions.The FES parameters were generated using the frequency and pulse width dual regulation algorithm based on the time domain feature proposed by Dr.Zhou.4.Design of the wearable 4-channel FES module: Firstly,a BLE module circuit was designed to transmit the recognition results and the FES parameters.Then,a wearable 4-channel FES stimulator was designed based on the complementary current source and time-division multiplexing method proposed by Dr.Wang[49]? The module generates a FES pulse waveform conducted to the corresponding stimulation channel according to the FES parameter.5.Joint debug of wearable EMGB system and healthy volunteer experiment verification: The s EMG detecting armband,recognition module and FES module were jointly debugged.A preliminary healthy volunteer experiment was designed to validate the effects of the wearable EMGB system.The experimental results proved that the wearable EMGB system can achieve four kinds of rehabilitation training actions(grasping,wrist flexion,wrist extension and five-finger extension)with the control of healthy limbs.