Sleep monitoring method for processing brain wave signals based on TB-TF-BiGRU model
The invention relates to a sleep monitoring method for processing brain wave signals based on a TB-TF-BiGRU model, and the method comprises the following steps: 1, obtaining collected brain wave initial signals, and processing and packaging brain wave signal data; 2, the electroencephalogram signal...
Saved in:
Main Authors | , , |
---|---|
Format | Patent |
Language | Chinese English |
Published |
01.11.2022
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | The invention relates to a sleep monitoring method for processing brain wave signals based on a TB-TF-BiGRU model, and the method comprises the following steps: 1, obtaining collected brain wave initial signals, and processing and packaging brain wave signal data; 2, the electroencephalogram signal data are processed through a TB-TF-BiGRU model, and sleep stages are obtained; and step 3, processing the feature vector through a Softmax classifier to complete sleep whole process monitoring. According to the method, the effective utilization rate of the initial brain wave signal is improved through resampling and difference reduction processing of the brain wave signal, and the difference of individual data is reduced; a three-branch time-frequency bidirectional model and a gated recurrent neural network are used for extracting different time-domain information from time-domain signals of electroencephalogram through different convolution kernels, so that the accuracy of sleep staging is improved; the brain wave |
---|---|
Bibliography: | Application Number: CN202210899835 |