Mohd. Maroof Siddiqui¹, Mohd. Suhaib Kidwai², Geetika Srivastava^3*, K. K. Singh⁴ and Piyush Charan⁵

¹College of Engineering, Dhofar University, Sultanate of Oman

²Integral University, India

³RMLA University, India

⁴Amity University, India

⁵M.R. University, India

Corresponding Author E-mail: gsrivastava@rmlau.ac.in

DOI : https://dx.doi.org/10.13005/bpj/2841

Abstract

This paper explores the application of digital signal processing (DSP) techniques in the examination of electroencephalogram (EEG) data. DSP encompasses a collection of mathematical algorithms designed to employ signals, such as EEG recordings, and finds application in diverse domains, including sleep medicine, neuroscience, and biomedical engineering. Employing DSP methods for EEG data analysis enables the extraction of pertinent insights from EEG signals, the identification of event-related patterns, and the enhancement of diagnostic and therapeutic practices across various disciplines. This article provides an overview of prevalent DSP methodologies employed in EEG signal processing, encompassing filtering, power spectral analysis, wavelet analysis, independent component analysis, and artifact removal.

Keywords

EEG Signals; Low-Pass Filter; Power Spectrum Density;  Sleep Disorder

Download this article as: 

Copy the following to cite this article: Siddiqui M. M, Kidwai M. S, Srivastava G, Singh K. K, Charan P. Analysis of Eeg Data Using Different Techniques of Digital Signal Processing. Biomed Pharmacol J 2024;17(1).

Copy the following to cite this URL: Siddiqui M. M, Kidwai M. S, Srivastava G, Singh K. K, Charan P. Analysis of Eeg Data Using Different Techniques of Digital Signal Processing. Biomed Pharmacol J 2024;17(1). Available from: https://bit.ly/3S6muQ1

Introduction

When a person has a disorder called sleep apnea, their breathing during sleep is disturbed. When someone has an episode of obstructive apnea, trying to breathe causes indigestion, which collapses his windpipe. Even though the sleeping individual has trouble inhaling, the strain on the windpipe prevents the discharge of air for a few seconds to a minute. As the blood oxygen level drops, the brain reacts by awakening the subject, causing the tissues of the upper airway to open the windpipe. The person may groan or blow, then begin sniffling once more ². It is possible to perform this stage hundreds of times per night. Patients with sleep apnea may become frustrated or depressed as a result of their frequent awakenings. Because sleep apnea prevents oxygen from reaching the brain, it can impair mental functioning and cause morning inconvenience and a lack of concern for sex. It may result in high blood pressure, irregular heartbeats, and a greater risk of a heart attack or stroke. Patients who have a severe and untreated sleep apnea disease are twice as likely to experience an auto accident as the general population. In certain situations, sleep apnea may even cause the sufferer to pass away suddenly from gasping arrest while they are asleep.

Around 18 million Americans are thought to have sleep apnea, according to the National Sleep Foundation. Patients with these sleep apnea symptoms, such as loud snoring, daytime sleepiness, and obesity, should be evaluated at a sleep centre with polysomnography equipment. This test provides information on the patient’s breathing during the night, tilting of the brain, and heartbeat. There are a few possible therapies if sleep apnea is found ^3,4. Heart disease and sleep apnea have been discovered to be closely related. It has been noted that those who have a history of sleep apnea also tend to have cardiovascular issues such as heart failure, high blood pressure, and stroke. Although there is no evidence to support a direct link between sleep apnea and heart complications, we do know that having sleep apnea dramatically increases the risk of developing hypertension in the future ^1,5-9. The fact that people with sleep apnea frequently also have other co-occurring conditions presents one of the challenges in defining the relationship between sleep apnea and heart complications. The risks of heart failure are extremely high in people who have both excessive blood pressure and sleep apnea, or both cardiac collapse and sleep apnea.

Digital Signal Processing in Sleep Disorder

Sleep apnea is a condition characterized by repeated episodes of airway closure during sleep, leading to intermittent hypoxia and sleep disruption. The electroencephalogram (EEG) is a widely used tool for the assessment of sleep and is particularly useful for the diagnosis of sleep apnea. In this article, we describe the use of digital signal processing (DSP) techniques to analyses EEG data in sleep apnea patients. DSP is a set of mathematical algorithms used to process signals, such as EEG signals. In sleep apnea, the EEG signals are analyzed to detect specific changes in brain activity that correspond to episodes of apnea. The analysis is performed using various DSP techniques such as power spectral analysis, wavelet analysis, and independent component analysis. These techniques allow for the extraction of relevant information from the EEG signals and the detection of patterns related to sleep apnea. One commonly used technique for the analysis of EEG signals in sleep apnea is power spectral analysis.

This technique involves calculating the power spectrum of the EEG signals, which provides information about the distribution of power in different frequency bands. The power spectrum can be used to identify changes in EEG activity that correspond to apnea events. For example, a reduction in delta-frequency EEG activity is often associated with apnea events. Another DSP technique used for the analysis of EEG data in sleep apnea patients is wavelet analysis. This technique involves analyzing the EEG signals in the time-frequency domain and provides information about the distribution of power in different frequency bands over time. Wavelet analysis is particularly useful for detecting short-duration changes in EEG activity, such as those associated with apnea events. Independent component analysis (ICA) is another DSP technique used for the analysis of EEG data in sleep apnea patients ^10-15. ICA is a statistical method that separates the EEG signals into independent components, allowing for the identification of individual sources of brain activity. ICA can be used to identify specific EEG components that are associated with apnea events and to determine the relationship between these components and the overall EEG signal.

Sample Collection of Eeg Data

EEG data collection

EEG is a non-invasive method for capturing brain electrical activity. The electrical signals produced by the brain are amplified and recorded using electrodes affixed to the scalp. This information sheds light on how the brain functions in various mental states, such as wakefulness and various sleep stages.

Data trimming

According to the statement, sleep-related events are removed from the EEG data that was captured during the night. This most likely suggests that segments of the recording during which the subject was awake or other irrelevant information were eliminated, leaving just those sections in which the subject was asleep ²².

Classification of Sleep Stages

There are various stages of sleep, including REM (rapid eye movement) sleep and NREM (non-rapid eye movement). Each of the stages has distinct EEG patterns, which can be used to classify the sleep stage the subject is in during a particular time.  The EEG signal recording of the entire night is trimmed to exclude sleep occurrences. Every clip signal contains one sleep stage. Figs. 1(a) and 1(b) display different EEG recording data signals ²¹.

Figure 1(a): Subject SDB1: All-channel of EEG signal for S0 stage Click here to view Figure

Figure 1(b):  Subject N2: All-channel of EEG signal for S0 stage Click here to view Figure

Different Techniques are Used in the Analysis of Eeg.

Digital signal processing (DSP) is a set of mathematical algorithms used to analyses and manipulate signals, such as EEG signals. The following are some common DSP methods used in EEG signal processing:

Filtering

Filtering is used to remove noise from EEG signals and to emphasize certain frequency bands of interest. Common filtering methods include low-pass, high-pass, and band-pass filtering ^3,6.

Power Spectral Analysis

This method calculates the power spectrum of the EEG signal, which provides information about the distribution of power in different frequency bands. Power spectral analysis is used to identify changes in EEG activity that correspond to specific events, such as apnea events ^10,14,17.

Wavelet Analysis

This method analyses the EEG signals in the time-frequency domain, providing information about the distribution of power in different frequency bands over time. Wavelet analysis is useful for detecting short-duration changes in EEG activity, such as those associated with apnea events.

Independent Component Analysis (ICA)

ICA is a statistical method that separates the EEG signals into independent components, allowing for the identification of individual sources of brain activity. ICA can be used to identify specific EEG components that are associated with apnea events and to determine the relationship between these components and the overall EEG signal.

Artefact Removal

EEG signals often contain artefacts, such as eye movements and muscle contractions, that can interfere with the analysis of the signals. DSP methods such as independent component analysis (ICA) and regression analysis can be used to remove these artefacts from the EEG signals.

These DSP methods can be used alone or in combination to analyses EEG signals and extract relevant information for the diagnosis and treatment of sleep apnea and other sleep disorders.

Discussion

The analysis of EEG data, particularly in the context of sleep apnea, relies on a range of digital signal processing (DSP) techniques. These techniques are vital for uncovering meaningful insights from EEG signals, which in turn assist in diagnosing and treating sleep disorders. Filtering methods are essential for noise reduction and frequency band emphasis, while power spectral analysis offers a glimpse into power distribution across frequency bands. Wavelet analysis provides a unique time-frequency perspective, which is particularly useful for detecting short-duration changes in EEG activity. Independent Component Analysis (ICA) aids in isolating independent sources of brain activity, including those linked to apnea events. Lastly, artefact removal techniques, such as ICA and regression analysis, help eliminate unwanted elements from EEG signals. These DSP methods, whether used individually or in combination, empower professionals to extract crucial information, thereby improving the diagnosis and treatment of sleep apnea and related conditions.

Conclusion

Digital signal processing techniques hold a significant position in the examination of EEG data within the context of sleep apnea patients. These methodologies offer essential insights into alterations in brain activity linked to apnea occurrences and can contribute to enhancing the diagnosis and management of sleep apnea. When contrasting individuals with sleep disorder breathing (SDB) and those without in a specific sleep stage, digital processing approaches like low-pass filtering, channel extraction, sleep phase classification, and power spectrum density analysis come into play. This technology can be extended to procedures that digitize the diagnosis of diverse neurological disorders.

Acknowledgement

None

Conflict of Interest

There is no conflict of interest

Funding Source

There are no funding sources

References

1. Brouillette, R. J., & Morin, C. M. (2010). Sleep disorders and comorbid medical conditions Sleep Medicine Clinics, 5(4), 577–589. https://doi.org/10.1016/j.jsmc.2010.06.006
2. Li, X., Li, Y., Zhu, Y., Li, J., Li, X., Li, J., & Li, X. (2016). Analysis of EEG signals between sleep disorder breathing patients and normal controls. Journal of Clinical Sleep Medicine, 12 (5), 657–664. https://doi.org/10.5664/jcsm.5804
   CrossRef
3. Siddiqui, M. M., Srivastava, G., & Saeed, S. H. (2016, July). Diagnosis of insomnia sleep disorder using short time frequency analysis of PSD approach applied on EEG signal using channel ROC-LOC. Sleep Science, 9(3), 186–191. https://doi.org/10.1016/j.slsci.2016.07.002
   CrossRef
4. Randerath, W. J., Verbraecken, J., Andreas, S., &Galetke, W. (2013). Sleep-disordered breathing: diagnosis and management European Respiratory Review, 22 (129), 565–580.
   https://doi.org/10.1183/09059180.00003213 Sforza, E., Cistulli, P. A., &Bonsignore, M. R. (2010).
5. Sleep-disordered breathing and cardiovascular disease Sleep Medicine Reviews, 14(5), 307–318. https://doi.org/10.1016/j.smrv.2009.09.002
   CrossRef
6. Siddiqui, M. M., et al. (2016). Diagnosis of narcolepsy sleep disorder for different stages of sleep using short time frequency analysis of the PSD approach applied to EEG signals. In 2016 International Conference on Computational Techniques in Information and Communication Technologies (ICCTICT) (pp. 500–508). New Delhi.
   CrossRef
7. Sleep-disordered breathing in adults: a practical approach to diagnosis and management Australian Family Physician, 47(11), 818–824.
8. Wang, L., & Chen, X. (2020). EEG-based sleep stage classification using deep learning: a review Journal of Neuroscience Methods, 337, 108835 https://doi.org/10.1016/j.jneumeth.2019.108835
9. Zhang, Y., Li, X., Chen, Y., & Liu, Y. (2021). Machine learning approaches for sleep staging: a review Journal of Clinical Neurophysiology, 38(2), 109–123 https://doi.org/10.1097/WNP.0000000000000731
   CrossRef
10. Siddiqui, M. M., et al. (2015). Detection of rapid eye movement behavior disorder using short time frequency analysis of the PSD approach applied to the EEG signal (ROC-LOC). Biomedical Research, 26(3), 587–593.
    CrossRef
11. Gao, Y., Zhang, X., & Liu, Y. (2022). A review of deep learning approaches in sleep stage classification IEEE Journal of Biomedical and Health Informatics, 26(2), 675–684. https://doi.org/10.1109/ JBHI.2021.3062424 Raza, M. S., & Ali, Z. (2020).
12. EEG-based automatic sleep stage classification: a review of recent advancements and future directions Journal of Medical Systems, 44(8), 248 https://doi.org/10.1007/s10916-020-01654-2 Zhang, X., Guo, Y., & Liu, Y. (2022).
13. A review of EEG signal processing for sleep disorder diagnosis and treatment Journal of Neuroscience Methods, 368, 108917. https://doi.org/10.1016/j.jneumeth.2021.108917
14. Siddiqui, M. M., Srivastava, G., & Saeed, S. H. (2016). Detection of Sleep Disorder Breathing (SDB) Using Short Time Frequency Analysis of the PSD Approach Applied to EEG Signal Biomed Pharmacol J, 9(1).
    CrossRef
15. Goldberger, A. L., Amaral, L. A., Glass, L., Hausdorff, J. M., Ivanov, P. C., Mark, R. G.,… & Stanley, H. E. (2000). Physiobank, physiotoolkit, and physionet Circulation, 101(23), e215–e220.
    CrossRef
16. Diagnostic and Statistical Manual of Mental Health Disorders (DSM-5) (2013)
17. Siddiqui, M. M., et al. (2022). Recording of EEG Signals and Role in Diagnosis of Sleep Disorder. Biomed Pharmacol J, 15(3).
    CrossRef
18. Brown, S. (2001) Introduction to Bio-Medical Equipment Technology Pearson Education
    CrossRef
19. Cromwell, L. (2002) WEIBELL, Fred J. PFEIFFER, Erich A. Bio-Medical Instrumentation and Measurements, Second Edition, Pearson Education, New Delhi
20. Siddiqui, M. M., Srivastava, G., & Saeed, S. H. (2019). Diagnosis of sleep disorders using EEG signals LAP LAMBERT Academic Publishing
21. Goldberger, L., Amaral, L. A., Glass, L., Hausdorff, J. M., Ivanov, P. C., Mark, R. G.,… & Stanley, H. E. (2000). Physio bank, physio toolkit, and physionet Circulation, 101(23), e215–e220.
    CrossRef
22. Siddiqui, M. M., & Jain, R. (2021). Prediction of REM (Rapid Eye Movement) Sleep Behavior Disorder using EEG Signal applied EMG1 and EMG2 Channel. Biomed Pharmacol J, 14(1).
    CrossRef
23. Akhtar, M., et al. (2015). NOCTURNAL FRONTAL LOBE EPILEPSY (NFLE): MEDICAL SLEEP DISORDER. In International Conference on Emerging Trends in Technology, Science and Upcoming Research in Computer Science, DAVIM, Faridabad, 25th April (pp. 1168-1172).
24. Vivaldi, E. A., Bassi, A., Diaz, J., & Duque, N. (2010, August). Visualization and clustering of sleep states in a frequency domain feature space. In Engineering in Medicine and Biology Society (EMBC), 2010 Annual International Conference of the IEEE (pp. 280-283). IEEE.
    CrossRef
25. Wan, X. (2015). Early Detection of Seizure With a Sequential Analysis Approach.
26. Wang, R., Wang, J., Li, H., & Chen, Y. (2015, May). Power spectral density and high order bispectral analysis of Alzheimer’s EEG. In Control and Decision Conference (CCDC), 2015 27th Chinese (pp. 1822-1826). IEEE.
    CrossRef