Research Article

Automatic Sleep Stage Classification Based on Convolutional Neural Network and Fine-Grained Segments

Table 1

The AASM staging criteria.

Sleep stageClassification criteria

Wake (W)Most epochs in this stage consist of alpha (8–13 Hz) activity or low voltage, mixed (2–7 Hz) with frequency activity.
N1Most epochs in this stage have low-voltage, mixed (2–7 Hz) activity; small epoch contains alpha (8–13 Hz) activity. Slow rolling movement usually lasts several seconds often seen in early N1.
N2Typical features of this stage are the appearance of sleep spindles, complex, and some high voltage activities (<2 Hz).
N3Most epochs in this stage begin to show high-voltage (>75 μV) and low-frequency (<2 Hz) activity.
REMThis stage has a pretty active performance. It has relatively low-voltage mixed-frequency (2–7 Hz) EEG with episodic rapid eye movements and absent or reduced chin EMG activity.