Towards A Wearable Bci System For Stroke Rehabilitation

Brain-computer interfaces(BCI)can translate the neural activities of human brains into control commands,and provide communication capabilities for those people with neuromuscular diseases.Clinical studies have shown that BCI can help those stroke patients with active rehabilitation of motor function,which achieves better rehabilitation effects than traditional treatment methods.However,most of the current BCI research is still limited in laboratory exploration,and the the BCI system is expensive,bulky,and complicated to set up.To develop a wearable BCI system for community and home use,this thesis focuses on the integration of the EEG acquisition system,the embedded development,the GUI software development,and the preparation of new electrodes.The following work is carried out:(1)Based on the EEG amplifier JAGA16,we develop a wearable BCI system prototype for the stroke patients’ hand function rehabilitation.To verify the effectiveness of the prototype,the subjects are recruited for alpha wave test,steady-state visual evoked potentials(SSVEP)test,and motor imagery test.Prominent signal characteristics can be observed after analyzing the EEG signal,which verifies the performance of the prototype.(2)To upgrade the hardware and software of the embedded device in the prototype,an EEG signal amplifier is developed,which includes an RHD2216 signal amplification module,an MCU control module,a local data storage module,a bluetooth transmission module,and a power module.The amplifier improves the integration of the system.We then conduct a test of sinusoidal signal acquisition and a test of input reference noise to verify the performance of the amplifier.We also develop a GUI software with Python language to simplify the user manipulation.The software has functions of EEG signal reception,real-time signal display,experimental paradigms integration,and signal decoding.We then do tests on the software,which proves the correctness and reliability of the software.(3)To upgrade the wet gel electrodes used in the prototype,we design soft ionic-hydrogel based electrodes for EEG signal recording,which overcomes the shortcomings of wet gel electrodes with longer preparation time and the inconvenience to wash hair after experiments.The electrode leaves no residue on the scalps and it achieves a low skin-electrode contact impedance with unique structures.The experimental results show that the ionic-hydrogel electrodes have similar electrical performance,noise level and contact impedance with wet gel electrodes,significantly better than the dry and water-based electrodes.The EEG signal measurements and SSVEP experiments with five subjects show the potential of the ionic-hydrogel electrodes for BCI application.Finally,to verify the real-time performance of the system in stroke rehabilitation applications,we build an online experimental platform.We then recruit six healthy subjects and two stroke patients with right-handed hemiplegic to participate in the classic motor imagery experiments and verify the effectiveness of the system.