Enhancing Brain-Machine Interface Throughput Using Simultaneous Activation Detection

摘要

In this work, we investigate the viability of a novel combination of evoked responses as input signals for a general-purpose brain-machine interface (BMI). We demonstrate response accuracy to alphanumeric stimuli in valid and mirror-reversed orientations, and show task-related activity differences correlated with rotation degree and character validity in superior parietal and inferior frontal gyrus regions of the brain. By observing simultaneous task-related activation in spatially dissociated regions, we increase the amount of information used for inferring user intent in control interfaces.