Simple method for detecting sleep episodes in rats ECoG using machine learning

In this paper we propose a new method for the automatic recognition of state behavioral sleep (BS) and waking (WS) in freely moving rats using their electrocorticographic (ECoG) data. Three-channels ECoG signals were recorded from frontal left, right occipital cortical areas. We employed simple artificial neural network (ANN), which mean values standard deviations two or three channels used as inputs ANN. Results wavelet-based BS/WS same data to train ANN evaluate correctness our classifier. tested different combinations detecting BS/WS. Our results showed that accuracy classification did not depend on ECoG-channel. For any ECoG-channel, networks trained one rat applied another with an at least 80~\%. Itis important very topology achieve relatively high classification. classifier was based linear combination input some weights, these weights could be replaced by averaged all ANNs without decreases accuracy. all, introduce does require additional training. It is enough know coefficients equations suggested paper. The proposed fast performance computations, therefore it real time experiments. might demand preclinical studies rodents vigilance control monitoring sleep-wake patterns.