In Clinical Practice the Most Common Indica- Tions for Polysomnography (psg) Are Investi- Gation and Treatment of Obstructive Sleep

73 EEG Electrode Placement in Polysomnography—Ruehland et al IN CLINICAL PRACTICE THE MOST COMMON INDICATIONS FOR POLYSOMNOGRAPHY (PSG) ARE INVESTIGATION AND TREATMENT OF OBSTRUCTIVE SLEEP apnea (OSA). For OSA diagnosis the main outcome measures from PSG are: (1) the apnea-hypopnea index (AHI), which is a measure of sleep disordered breathing events (apneas and hypopneas) per hour of sleep, (2) the arousal index (ArI) which is a measure of sleep disruption per hour of sleep and (3) various sleep scoring summary statistics describing sleep quality or sleep architecture (e.g., sleep efficiency). For these measures, scoring of arousals and sleep are not only important in their own right, but they also impact on other measures. For example, total sleep time (TST) is used as the denominator for both the AHI and ArI, and some criteria allow for hypopnea scoring if airflow reduction is accompanied by a cortical arousal.1 The scoring of sleep and arousals relies on visual inspection of continuous surface electroencephalography (EEG), electromyography (EMG), and electrooculography (EOG) measurements. Rechtschaffen and Kales (R&K),2 the first consensus-based guidelines for scoring of sleep, recommended recording a minimum of one channel of central EEG (either C3/A2 or C4/A1) during PSG. The one-derivation minimum was recommended due to device limitations and because it was thought that regional differences in scalp areas were not critical for sleep scoring. Subsequently, other authors have suggested that regional differences may be important3 and have recommended the use of more than one EEG derivation.4 In agreement with this view the 2007 AASM Manual for the Scoring of Sleep and Associated Events1 recently recommended the use of 3 standard EEG derivations for scoring of sleep; including frontal, central, and occipital derivations. The evidence review paper underpinning the AASM manual,5 stated the recommendations follow from the current less restrictive device limitations and from observations in healthy subjects that, although sleep spindles may be generally recorded optimally over central regions,3,6-8 this is not the case for other ELECTRODE PLACEMENT, SLEEP AND CORTICAL AROUSAL SCORING