Open AccessJoint sleep staging model based on pressure-sensitive sleep signal
Author(s) -
Yudong Huang,
Liwei Liang
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/740/1/012159
Subject(s) - sleep (system call) , sleep stages , computer science , artificial intelligence , stage (stratigraphy) , signal (programming language) , pattern recognition (psychology) , medicine , polysomnography , electroencephalography , operating system , psychiatry , paleontology , programming language , biology
Sleep quality is an important indicator of human health, and sleep staging is a prerequisite for assessing sleep quality. In this paper, the CNN-BILSTM sleep staging model is proposed based on the pressure-sensing sleep signal collected by the smart mattress. The CNN network is used as the feature extraction part for the problem of the traditional sleep staging model in the pressure-sensing sleep stage [2]. This part can automatically extract the pressure sense. The staging characteristics of the sleep signal, combined with the BiLSTM model for sleep staging. In the experiment, the sleep data of 11 different experimenters based on smart mattress collection were used to identify the three sleep stages, in order to verify the superiority of the sleep stage of the model [3]. Two kinds of contrast experiments were used, one was compared with a single CNN stage, the BiLSTM model was compared, and the other was compared with the traditional feature extraction based sleep stage model. The experimental results show that the model has an accuracy of more than 80%. Better than traditional sleep staging and single deep learning sleep staging program.