Waking EEG functional connectivity in middle-aged and older adults with obstructive sleep apnea

Sleep Apnea Detection Using EEG: A Systematic Review of Datasets, Methods, Challenges, and Future Directions

PURPOSE

Sleep Apnea (SA) affects an estimated 936 million adults globally, posing a significant public health concern. The gold standard for diagnosing SA, polysomnography, is costly and uncomfortable. Electroencephalogram (EEG)-based SA detection is promising due to its ability to capture distinctive sleep stage-related characteristics across different sub-band frequencies. This study aims to review and analyze research from the past decade on the potential of EEG signals in SA detection and classification focusing on various deep learning and machine learning techniques, including signal decomposition, feature extraction, feature selection, and classification methodologies.

METHOD

A systematic literature review using the preferred reporting items for systematic reviews and meta-Analysis (PRISMA) and PICO guidelines was conducted across 5 databases for publications from January 2010 to December 2024.

RESULTS

The review involved screening a total of 402 papers, with 63 selected for in-depth analysis to provide valuable insights into the application of EEG signals for SA detection. The findings underscore the potential of EEG-based methods in improving SA diagnosis.

CONCLUSION

This study provides valuable insights, showcasing significant advancements while identifying key areas for further exploration, thereby laying a strong foundation for future research in EEG-based SA detection.

Changes in the Spatial Structure of Synchronization Connections in EEG During Nocturnal Sleep Apnea

This study involved 72 volunteers divided into two groups according to the apnea-hypopnea index (AHI): AHI>15 episodes per hour (ep/h) (main group, n=39, including 28 men, median AHI 44.15, median age 47), 0≤AHI≤15ep/h (control group, n=33, including 12 men, median AHI 2, median age 28). Each participant underwent polysomnography with a recording of 19 EEG channels. Based on wavelet bicoherence (WB), the magnitude of connectivity between all pairs of EEG channels in six bands was estimated: Df1 0.25;1, Df2 1;4, Df3 4;8, Df4 8;12, Df5 12;20, Df6 20;30 Hz. In all six bands considered, we noted a significant decrease in symmetrical interhemispheric connections in OSA patients. Also, in the main group for slow oscillatory activity Df1 and Df2, we observe a decrease in connection values in the EEG channels associated with the central interhemispheric sulcus. In addition, patients with AHI>15 show an increase in intrahemispheric connectivity, in particular, forming a left hemisphere high-degree synchronization node (connections PzT3, PzF3, PzFp1) in the Df2 band. When considering high-frequency EEG oscillations, connectivity in OSA patients again shows a significant increase within the cerebral hemispheres. The revealed differences in functional connectivity in patients with different levels of AHI are quite stable, remaining when averaging the full nocturnal EEG recording, including both the entire sleep duration and night awakenings. The increase in the number of hypoxia episodes correlates with the violation of the symmetry of interhemispheric functional connections. Maximum absolute values of correlation between the apnea-hypopnea index, AHI, and the WB synchronization strength are observed for the Df2 band in symmetrical EEG channels C3C4 (-0.81) and P3P4 (-0.77). The conducted studies demonstrate the possibility of developing diagnostic systems for obstructive sleep apnea syndrome without using signals from the cardiovascular system and respiratory activity.

Phase-Amplitude Coupling in Theta and Beta Bands: A Potential Electrophysiological Marker for Obstructive Sleep Apnea

Background

Phase-amplitude coupling (PAC) between the phase of low-frequency signals and the amplitude of high-frequency activities plays many physiological roles and is involved in the pathological processed of various neurological disorders. However, how low-frequency and high-frequency neural oscillations or information synchronization activities change under chronic central hypoxia in OSA patients and whether these changes are closely associated with OSA remains largely unexplored. This study arm to elucidate the long-term consequences of OSA-related oxygen deprivation on central nervous system function.

Methods

: We screened 521 patients who were clinically suspected of having OSA at our neurology and sleep centers. Through polysomnography (PSG) and other clinical examinations, 103 patients were ultimately included in the study and classified into mild, moderate, and severe OSA groups based on the severity of hypoxia determined by PSG. We utilized the phase-amplitude coupling (PAC) method to analyze the modulation index (MI) trends between different frequency bands during NREM (N1/N2/N3), REM, and wakefulness stages in OSA patients with varying severity levels. We also examined the correlation between the MI index and OSA hypoxia indices.

Results

Apart from reduced N2 sleep duration and increased microarousal index, the sleep architecture remained largely unchanged among OSA patients with varying severity levels. Compared to the mild OSA group, patients with moderate and severe OSA exhibited higher MI values of PAC in the low-frequency theta phase and high-frequency beta amplitude in the frontal and occipital regions during N1 sleep and wakefulness. No significant differences in the MI of phase-amplitude coupling were observed during N2/3 and REM sleep. Moreover, the MI of phase-amplitude coupling in theta and beta bands positively correlated with hypoxia-related indices, including the apnea-hypopnea index (AHI) and oxygenation desaturation index (ODI), and the percentage of oxygen saturation below 90% (SaO2<90%).

Conclusion

OSA patients demonstrated increased MI values of theta phase and beta amplitude in the frontal and occipital regions during N1 sleep and wakefulness. This suggests that cortical coupling is prevalent and exhibits sleep-stage-specific patterns in OSA. Theta-beta PAC during N1 and wakefulness was positively correlated with hypoxia-related indices, suggesting a potential relationship between these neural oscillations and OSA severity. The present study provides new insights into the relationship between neural oscillations and respiratory hypoxia in OSA patients.

Compensatory mechanisms of reduced interhemispheric EEG connectivity during sleep in patients with apnea

We performed a mathematical analysis of functional connectivity in electroencephalography (EEG) of patients with obstructive sleep apnea (OSA) (N = 10; age: 52.8 ± 13 years; median age: 49 years; male/female ratio: 7/3), compared with a group of apparently healthy participants (N = 15; age: 51.5 ± 29.5 years; median age: 42 years; male/female ratio: 8/7), based on the calculation of wavelet bicoherence from nighttime polysomnograms. Having observed the previously known phenomenon of interhemispheric synchronization deterioration, we demonstrated a compensatory increase in intrahemispheric connectivity, as well as a slight increase in the connectivity of the central and occipital areas for high-frequency EEG activity. Significant changes in functional connectivity were extremely stable in groups of apparently healthy participants and OSA patients, maintaining the overall pattern when comparing different recording nights and various sleep stages. The maximum variability of the connectivity was observed at fast oscillatory processes during REM sleep. The possibility of observing some changes in functional connectivity of brain activity in OSA patients in a state of passive wakefulness opens up prospects for further research. Developing the methods of hypnogram evaluation that are independent of functional connectivity may be useful for implementing a medical decision support system.

Associations between cortical activation and network interaction during sleep

Cortical activation and network interaction, two characterizations of the cortical states, are separately studied in most previous studies. To further clarify the underlying mechanism, the association between these two indicators during sleep was investigated in this study. Twenty healthy individuals were enrolled and all of them underwent overnight polysomnography (PSG) recording. The relative spectral powers and the phase transfer entropy (PTE) of various frequency components were extracted from 6 electroencephalographic (EEG) channels, to assess the cortical activation and network interaction, respectively. Pearson correlation coefficient was employed to estimate their associations. The results suggested that there was a negative correlation between spectral power and phase transfer entropy in δ and α frequency bands during sleep. As the sleep deepened, an increased negative correlation in the δ frequency band was noted, but the negative correlation became less extreme in the α frequency band. The extremum of the correlation coefficient was noted in δ of N3, and α of Wake. Overall, this study provides a connection between these two cortical activity assessments, especially reveals the variable characteristics of different frequency components, which is conducive to better understand sleep state.