Table 1. Overview of sleep scoring methodologies.
Reference | Motor parameters | Brain activity parameters | Manual calibration |
---|---|---|---|
Brankack and colleagues, 2010 | EMG | EEG power: delta (1–4 Hz), theta2 (7–8.5 Hz), and gamma2 (52–70 Hz) | Manual scoring of 5% of data to train LDA or classification tree |
Crisler and colleagues, 2008 | EMG | EEG: 107 temporal and spectral parameters | Manual scoring of 2 h of data to train SVM |
Gross and colleagues, 2009 | EMG | EEG: Delta, theta, sigma, and beta bands | Manual determination of thresholds |
Louis and colleagues, 2004 | EMG | EEG: Delta, theta, alpha, beta, and gamma | Manual scoring of 1 d of data to establish thresholds |
Rytkönen and colleagues, 2011 | EMG | EEG: Delta, theta, alpha, beta, and gamma | Manual scoring of 5% of data to train naïve Bayes classifier |
Veasey and colleagues, 2000 | EMG | EEG: Delta, sigma, theta | Manual determination of thresholds |
Zeng and colleagues, 2012 | Doppler and video recording of movement and breathing | none | Manual scoring of 10% data to train SVM |
Liang and colleagues, 2012 | EMG | 13 EEG features | Manual scoring of 2 animals to establish thresholds for all others |
Stephenson and colleagues, 2009 | EMG | EEG: delta, alpha, theta, beta, and gamma power | No manual steps |
Abbreviations: EEG, electroencephalography; EMG, electromyography; LDA, linear discriminant analysis; SVM, support vector machine.