IM-tention: A Software for Brain-Computer Interface with Motor Recovery Purposes

The use of brain-computer interfaces (BCI) for motor recovery of people with disabilities has exponentially grown. In a motor recovery BCI, the user performs or attempts to perform a motor activity that produces changes in the electroencephalography (EEG) signal, such as the event-related desynchronization. Most of the developments in the BCI area were undertaken with software under license restrictions, which limits the design options and the changes necessary for clinical translation. For this reason, this article describes the design and performance evaluation of IM-tention, a custom software intended for a BCI for motor recovery. IM-tention was developed in Python and consists of a graphical user interface, a patient database, and the following blocks: Amplifier Communication, Signal Preprocessing, Feature Extraction, and Classifier Training (Calibration stage) or Feature Classification (Closed-loop stage). IM-tention supports the recording of EEG signals using the open source biopotential amplifiers OpenBCI and BioAmp. It detects, in real time, the movement or the motor intention of the user to activate an Application Device, such as a visual animation that closes the loop with a user’s perceivable feedback. For the performance evaluation, a BCI based on IM-tention was implemented. Six healthy participants used the BCI in a single session of foot dorsiflexion. The median Accuracy was 94% for the Calibration stage, and the median True Positive Rate was 70% for the Closed-loop stage. This could suggest that BCI based on IM-tention is capable of detecting the foot dorsiflexion. The next steps will be directed towards its use with patients with motor sequelae.