Minimally Invasive and Low-Cost BCI System Interprets the Will of the Subject by means of an Artificial Neural Network

The advances in BCI technology are evident and continuous, but the problems of portability and invasiveness of the devices that the subject must wear still remain to be solved. The paper presents a light, low-cost and non-invasive system that allows the disabled subject unable to communicate to interact with the environment and other people. Signals are acquired from a sensor applied to the finger, then are processed by an Artificial Neural Network. This allows the tetraplegic and anarthric subject to express their will and make choices through a graphic interface. The system can be greatly improved both in the quality and quantity of different signals acquired and in their simultaneous processing to discriminate different mental states, so as to improve communication, access to computer and home automation functions.