Georgia Tech researchers have created a new brain-machine interface (BMI) that uses a new class of nano-membrane electrode combined with flexible electronics and a new deep learning algorithm to help disabled people wirelessly control devices. The BMI could be used to control things like an electric wheelchair, interact with a computer, or operate a small robotic vehicle without having to use a hair-electrode cap or mess with wires.
The fully-portable BMI could also offer an improvement compared to conventional EEG. An EEG allows the measurement of signals from visually evoked potentials in the human brain. So far, the new BMI has been evaluated for measuring EEG signals with six human subjects but hasn’t been studied with disabled individuals.
The researchers say that BMI is an essential part of rehabilitation technology that allows those with ALS, chronic stroke, or other severe motor disabilities to control their prosthetic systems. Currently gathering the steady-state virtually evoked potentials, the signals from the brain, involves the patient wearing an electrode-studded hair cap with the electrodes, adhesives, and lots of wire.
The new class of flexible, wireless sensors and electronics that can be applied easily to the skin are used in the new BMI system. The flexible sensors send data to the telemetry unit using Bluetooth. The recorded EEG data from the brain is processed via a flexible circuit and sent to a computer to interpret the signals. The system has a wireless range of 15 meters.
The system uses three elastomeric scalp electrodes that are held on the head with a fabric band. The sensors are also placed on the neck and in an electrode placed below the ear. The electrodes are dry and soft, with no need for adhesive or gel. The next step in the research is to use the system on motor-impaired individuals.