Development of a bioimpedance-based human machine interface for wheelchair control

We present a new method, which is based on bioelectrical impedance of the trapezius muscles, to control wheelchair for the disabled people and the elderly. In our application, three electrodes were used for detecting the changes in movements of left and right trapezius muscles. The modified Howland current bridge supplies the 0.5 mA ac current for the measure system at the frequency of 50 kHz. NI PCI-6250 DAQ board were adopted to collect the data and Labview8.2 was used to implement the control system. The threshold value in detection algorithms applied in the system is automatically adjusted to the change in the measured signal magnitude. Pump value detection is used to detect an unexpected large change of the signal to avoid the wrong operation. As a result, we can find that the change of the signal according to the movement of the shoulder is very stable. Additionally, after some signal processing we can use shoulder movement to control LED on Labview8.2 with an accuracy of 100%.