Brain-machine interfaces: Principles and clinical application

The brain-machine interface (BMI) enables us to control machines and to communicate with others, not with the use of input devices, but through the direct use of brain signals. BMIs are classified into two types: the invasive type, which uses intracranial electrodes, and the noninvasive type, which uses skin electrodes or near infrared spectroscopy. The invasive type is further subdivided into two subtypes: a less invasive type, which uses brain surface electrodes and a highly invasive type, which uses needle microelectrodes. Noninvasive BMIs are promising for neurorehabilitation, while invasive BMIs are promising for use as neural prostheses for severely disabled people. The highly invasive type is characterized by high performance utilizing its detailed neural information, while the less invasive type is characterized by high feasibility for clinical application based on long-term stability. A fully-implantable wireless system is indispensable for the clinical application of invasive BMIs as it not only reduces the risk of infection but it also contributes to improving convenience.