Determination of Sleep Stage Separation Ability of Features Extracted from EEG Signals Using Principle Component Analysis

Vural, Cabir; Yıldız, Murat

doi:10.1007/s10916-008-9218-9

Cited by 44 publications

(19 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Significant features are fed to a Gaussian mixture model classifier for the classification of sleep stages and a classification accuracy of 88.7% is obtained. Vural and Yildiz [95] used the principal component analysis [96] for the classification of hybrid features and reported an accuracy of 69.98%. Langkvist et al [97] performed sleep stage classification using deep belief nets, an unsupervised feature learning approach.…”

Section: Discussionmentioning

confidence: 99%

Nonlinear Dynamics Measures for Automated EEG-Based Sleep Stage Detection

et al. 2015

View full text Add to dashboard Cite

Background: The brain's continuous neural activity during sleep can be monitored by electroencephalogram (EEG) signals. The EEG wave pattern and frequency vary during five stages of sleep. These subtle variations in sleep EEG signals cannot be easily detected through visual inspection. Summary: A range of time, frequency, time-frequency and nonlinear analysis methods can be applied to understand the complex physiological signals and their chaotic behavior. This paper presents a comprehensive comparative review and analysis of 29 nonlinear dynamics measures for EEG-based sleep stage detection. Key Messages: The characteristic ranges of these features are reported for the five different sleep stages. All nonlinear measures produce clinically significant results, that is, they can discriminate the individual sleep stages. Feature ranking based on the statistical F-value, however, shows that the third order cumulant of higher order spectra yields the most discriminative result. The distinct value ranges for each sleep stage and the discriminative power of the features can be used for sleep disorder diagnosis, treatment monitoring, and drug efficacy assessment.

show abstract

Section: Discussionmentioning

confidence: 99%

Nonlinear Dynamics Measures for Automated EEG-Based Sleep Stage Detection

et al. 2015

View full text Add to dashboard Cite

show abstract

“…The relative power was calculated by dividing the absolute power in each frequency band with the integrated power in the range 1–45 Hz. The central frequency, f c , and bandwidth, f σ 60 , were computed according to these formulas: where f L and f H represent the lowest and highest frequency that defines a given frequency band, and P ( H ) denotes the power at frequency f . Thus, the central frequency biomarker provides information about where the power is concentrated in a given frequency band, whereas the bandwidth provides information about how much the power is spread out around the central frequency.…”

Section: Methodsmentioning

confidence: 99%

EEG machine learning for accurate detection of cholinergic intervention and Alzheimer’s disease

Simpraga

Álvarez-Jiménez

Mansvelder

et al. 2017

Sci Rep

View full text Add to dashboard Cite

Monitoring effects of disease or therapeutic intervention on brain function is increasingly important for clinical trials, albeit hampered by inter-individual variability and subtle effects. Here, we apply complementary biomarker algorithms to electroencephalography (EEG) recordings to capture the brain’s multi-faceted signature of disease or pharmacological intervention and use machine learning to improve classification performance. Using data from healthy subjects receiving scopolamine we developed an index of the muscarinic acetylcholine receptor antagonist (mAChR) consisting of 14 EEG biomarkers. This mAChR index yielded higher classification performance than any single EEG biomarker with cross-validated accuracy, sensitivity, specificity and precision ranging from 88–92%. The mAChR index also discriminated healthy elderly from patients with Alzheimer’s disease (AD); however, an index optimized for AD pathophysiology provided a better classification. We conclude that integrating multiple EEG biomarkers can enhance the accuracy of identifying disease or drug interventions, which is essential for clinical trials.

show abstract

“…The instantaneous phase was also extracted using Hilbert transform, and the 95th percentile duration and size of the stable phase bursts (a phase bursts is defined as the period between phase slips) were calculated. In addition, we computed for all frequency bands and individualized frequency bands, defined as Alpha1 (APF = individually defined Alpha peak frequency): (APF–4 to APF–2) Hz, Alpha2: (APF–2 to APF) Hz, Alpha3: (APF to APF+2) Hz; Beta: (APF+2 to 30) Hz (Klimesch, 1999), 7 biomarkers: absolute, relative power, and power ratios, furthermore, the central frequency, power in central frequency, bandwidth and spectral edge (Vural and Yildiz, 2010; O'Gorman et al, 2013). In total, we extracted 177 biomarker values from each EEG trace (Table 1).…”

Section: Methodsmentioning

confidence: 99%

Integrative EEG biomarkers predict progression to Alzheimer's disease at the MCI stage

Poil¹,

Haan²,

Flier³

et al. 2013

Front. Aging Neurosci.

169

114

View full text Add to dashboard Cite

Alzheimer's disease (AD) is a devastating disorder of increasing prevalence in modern society. Mild cognitive impairment (MCI) is considered a transitional stage between normal aging and AD; however, not all subjects with MCI progress to AD. Prediction of conversion to AD at an early stage would enable an earlier, and potentially more effective, treatment of AD. Electroencephalography (EEG) biomarkers would provide a non-invasive and relatively cheap screening tool to predict conversion to AD; however, traditional EEG biomarkers have not been considered accurate enough to be useful in clinical practice. Here, we aim to combine the information from multiple EEG biomarkers into a diagnostic classification index in order to improve the accuracy of predicting conversion from MCI to AD within a 2-year period. We followed 86 patients initially diagnosed with MCI for 2 years during which 25 patients converted to AD. We show that multiple EEG biomarkers mainly related to activity in the beta-frequency range (13–30 Hz) can predict conversion from MCI to AD. Importantly, by integrating six EEG biomarkers into a diagnostic index using logistic regression the prediction improved compared with the classification using the individual biomarkers, with a sensitivity of 88% and specificity of 82%, compared with a sensitivity of 64% and specificity of 62% of the best individual biomarker in this index. In order to identify this diagnostic index we developed a data mining approach implemented in the Neurophysiological Biomarker Toolbox (http://www.nbtwiki.net/). We suggest that this approach can be used to identify optimal combinations of biomarkers (integrative biomarkers) also in other modalities. Potentially, these integrative biomarkers could be more sensitive to disease progression and response to therapeutic intervention.

show abstract

Determination of Sleep Stage Separation Ability of Features Extracted from EEG Signals Using Principle Component Analysis

Cited by 44 publications

References 9 publications

Nonlinear Dynamics Measures for Automated EEG-Based Sleep Stage Detection

Nonlinear Dynamics Measures for Automated EEG-Based Sleep Stage Detection

EEG machine learning for accurate detection of cholinergic intervention and Alzheimer’s disease

Integrative EEG biomarkers predict progression to Alzheimer's disease at the MCI stage

Contact Info

Product

Resources

About