Brain-computer interfaces: communication and restoration of movement in paralysis

摘要: The review describes the status of brain–computer or brain–machine interface research. We focus on non-invasive interfaces (BCIs) and their clinical utility for direct brain communication in paralysis motor restoration stroke. A large gap between promises invasive animal human BCI preparations reality characterizes literature: while intact monkeys learn to execute more less complex upper limb movements with spike patterns from regions alone without concomitant peripheral activity usually after extensive training, applications diseases such as amyotrophic lateral sclerosis stroke spinal cord lesions show only limited success, exception verbal paralysed locked-in patients. BCIs based electroencephalographic potentials oscillations are ready undergo studies commercial production an adjunct a major assisted device However, attempts train completely patients entering complete state no remaining eye movement failed. propose that lack contingencies goal directed thoughts intentions may be at heart this problem. Experiments chronically curarized rats support our hypothesis; operant conditioning voluntary control autonomic physiological functions turned out impossible preparation. In addition communication, consisting learning EEG slow cortical sensorimotor rhythm were demonstrated successful drug resistant focal epilepsy attention deficit disorder. First using MEG hand chronic single cases high some promise, but need evaluation well-controlled experiments. Invasive BMIs neuronal patterns, local field electrocorticogram constitute strategy choice severe paralysis. Future directions research should include regulation metabolism blood flow electrical magnetic stimulation (invasive non-invasive). series BOLD response functional resonance imaging (fMRI) near infrared spectroscopy tight correlation changes behaviour.