Abstract
For patients with ‘locked-in’ syndrome (LIS), brain-computer interfaces (BCIs) based on hemodynamics are a promising tool to regain communication. Visual and motor imagery are often-used information-encoding strategies despite frequent visuo-motor impairments in LIS. However, evoking imagery not grounded in recent sensory experience is very challenging. Therefore, somatosensory imagery might be more suitable, as somatosensation is preserved even in late-stage LIS. Using 7T fMRI in healthy subjects, we find that somatosensory imagery is associated with topographic decoding-weight patterns in primary somatosensory cortex and show high-accuracy decoding with little pre-training. We immediately implement these novel findings to drive brain-based communication, providing a promising proof-of-concept for a novel type of communication BCI with high clinical potential. Our new BCI-control strategy paves the way for several future applications. Next to brain-based communication, it might be employed for neurofeedback-based training to restore somatotopy in stroke and chronic pain patients suffering somatosensory deficits.
