Blinking Artefact Recognition in EEG Signal Using Artificial Neural Network

Artefacts are noises introduced to an EEG signal by patient’s movements and sources of electric field outside the patient’s body. The artefacts impede a doctor’s expertise and an automatic analysis of the signal. The most common and characteristic kind of artefacts are blinking artefacts. This paper presents a neural based approach to finding artefacts in the signal. The inputs to the network are different coefficients computed for a window of the signal, expressing some characteristic properties of blinking artefacts. 41 coefficients were designed. Sensitivity and correlation analyses were used to choose 14 coefficients for the network’s inputs. One used a large training set including coefficients for over 27000 windows, containing different kinds of EEG waves. Three classification algorithms were compared: k-neighbours, RBF networks and back propagation networks. The program was tested on the EEG signal and was highly evaluated by a domain expert.