Optimizing the individual differences of EEG signals through BP neural network algorithm for a BCI dialing system

Abstract

Brain-computer interface (BCI) establishes an additional pathway between brain and the external environment. With BCIs, paraplegic patients or the elderly people can communicate with others conveniently and finish some simple tasks individually. In this paper, we build an online brain computer interface system. The system consists of three main modules: electroencephalography (EEG) acquisition module, signal processing module and dialing system on the Android Platform. The system has several advantages, such as non-invasive, real-time, without training and the adaptability to different users by using backpropagation (BP) neural network. Experimental results show that by using BP neural network, the accuracy of the dialing system is improved. © Springer International Publishing 2013.