Reliable decoding of motor state transitions during imagined movement

Current non-invasive Brain Machine interfaces commonly rely on the decoding of sustained motor imagery activity. This approach enables a user to control brain-actuated devices by triggering predetermined motor actions. However, despite of its broad range of applications, this paradigm has failed so far to allow a natural and reliable control. As an alternative approach, we investigated the decoding of states transitions of an imagined movement, i.e. rest-to-movement (onset) and movement-to-rest (offset). We show that both transitions can be reliably decoded with accuracies of 71.47% for the onset and 73.31% for the offset (N = 9 subjects). Importantly, these transitions exhibit different neural patterns and need to be decoded independently. Our results indicate that both decoders are able to capture the brain dynamics during imagined movements and that their combined use could provide benefits in terms of accuracy and time precision.