A method is proposed for automatically choosing independent components (ICs) of interest from neonatal EEG data, with the aim of using them in further analysis to detect seizures. This procedure greatly reduces the amount of information which needs to be processed in the seizure detection system, and reduces the effect of noise and artefacts on its performance. The Fast ICA algorithm is used to generate the ICs, and the complexity of each IC is examined to determine those of interest. The Singular Value Fraction (SVF) measure is used to reduce the number of sources containing artefacts chosen. In the best case, the 12 channel EEG used in these tests is reduced to 2 or 3 sources of interest. In every case, at least 3 sources were removed that consisted of noise.