Endogenous brain–machine interface based on the correlation of EEG maps

In this paper, a non-invasive endogenous brain–machine interface (BMI) based on the correlation of EEG maps has been developed to work in real-time applications. The classifier is able to detect two mental tasks related to motor imagery with good success rates and stability. The BMI has been tested with four able-bodied volunteers. First, the users performed a training with visual feedback to adjust the classifier. Afterwards, the users carried out several trajectories in a visual interface controlling the cursor position with the BMI. In these tests, score and accuracy were measured. The results showed that the participants were able to follow the targets during the performed trajectory, proving that the EEG mapping correlation classifier is ready to work in more complex real-time applications aimed at helping people with a severe disability in their daily life.