Abstract:Objective.Sleep stage recognition has essential clinical value for evaluating human physical/mental condition and diagnosing sleep-related diseases. To conduct a five-class (wake, N1, N2, N3 and rapid eye movement) sleep staging task, twenty subjects with recorded six-channel electroencephalography (EEG) signals from the ISRUC-SLEEP dataset is used.Approach.Unlike the exist methods ignoring the channel coupling relationship and non-stationarity characteristics, we developed a brain functional connectivity method to provide a new insight for multi-channel analysis. Furthermore, we investigated three frequency-domain features: two functional connectivity estimations, i.e. synchronization likelihood (SL) and wavelet-based correlation (WC) among four frequency bands, and energy ratio (ER) related to six frequency bands, respectively. Then, the Gaussian support vector machine (SVM) method was used to predict the five sleep stages. The performance of the applied features is evaluated in both subject dependence experiment by ten-fold cross validation and subject independence experiment by leave-one-subject-out cross-validation, respectively.Main results.In subject dependence experiment, the results showed that the fused feature (fusion of SL, WC and ER features) contributes significant gain the performance of SVM classifier, where the mean of classification accuracy can achieve 83.97% ± 1.04%. However, in subject-independence experiment, the individual differences EEG patterns across subjects leads to inferior accuracy. Five typical domain adaptation (DA) methods were applied to reduce the discrepancy of feature distributions by selecting the optimal subspace dimension. Results showed that four DA methods can significantly improve the mean accuracy by 1.89%-5.22% compared to the baseline accuracy 57.44% in leave-one-subject-out cross-validation.Significance.Compared with traditional time-frequency and nonlinear features, brain functional connectivity features can capture the correlation between different brain regions. For the individual EEG response differences, domain adaptation methods can transform features to improve the performance of sleep staging algorithms.

Test Time Adaptation for Cross-Domain Sleep Stage Classification

Cross-scenario automatic sleep stage classification using transfer learning and single-channel EEG

Unsupervised Domain Adaptation by Statistics Alignment for Deep Sleep Staging Networks

ADAST: Attentive Cross-domain EEG-based Sleep Staging Framework with Iterative Self-Training

Deep adaptation network for subject-specific sleep stage classification based on a single-lead ECG

Day-Night Adaptation: An Innovative Source-free Adaptation Framework for Medical Image Segmentation

Sleep Quality Estimation with Adversarial Domain Adaptation: From Laboratory to Real Scenario

Automatic sleep staging of single-channel EEG based on domain adversarial neural networks and domain self-attention

Learning Instance-Specific Adaptation for Cross-Domain Segmentation

Online Test-Time Adaptation for Patient-Independent Seizure Prediction

DomainAdaptor: A Novel Approach to Test-time Adaptation

Generalizable Sleep Staging via Multi-Level Domain Alignment

Test-time adaptation for geospatial point cloud semantic segmentation with distinct domain shifts

Dual-Teacher Feature Distillation: A Transfer Learning Method for Insomniac PSG Staging

Multi-channel EEG-based sleep staging using brain functional connectivity and domain adaptation

Single-channel EEG sleep staging based on data augmentation and cross-subject discrepancy alleviation

ADATIME: A Benchmarking Suite for Domain Adaptation on Time Series Data

Online Adaptive Fault Diagnosis with Test-Time Domain Adaptation

Test-Time Adaptation for Depth Completion

Benchmarking EEG-based Cross-dataset Driver Drowsiness Recognition with Deep Transfer Learning

Multichannel Multidomain-Based Knowledge Distillation Algorithm for Sleep Staging with Single-Channel EEG