A study on recording and analysis conditions of polysomnography for automated computer analysis of sleep stages.

自動解析装置を用いて終夜睡眠ポリグラフを解析する際の記録条件と解析条件を検討した。対象は若年成人女性10名で,解析には0xford社製睡眠ポリグラフ記録･解析装置Medilog SAC 847 Systemを用い,第3夜の記録を視察判定と比較した｡徐波睡眠の指標となるδ波の最小振幅基準を初期設定の基線から±31/μVと,これより低い±25/μVに設定すると,徐波睡眠量,徐波睡眠率とも視察による判定とほぼ同じ解析結果が得られた。眼球運動の入力感度はREM睡眠量,REM睡眠率に影響しなかったが,オトガイ筋電図の入力感度を初期設定の1.5VにするとREM睡眠がほとんどみれなくなる例が多く,1.0VではREM睡眠量がこれまでの報告より少ないものの,視察判定に近い結果が得られた｡以上の研究結果から,我々が用いた自動解析装置では,オトガイ筋電図の入力感度を1.0V,δ波の最小振幅基準を±25～31/μVに設定すれば,視察判定とほぼ同じ解析結果が得られると考えられた｡Recording and analysis conditions of polysomnography were examined in 10 young adult female by using sleep analysis computer Medilog SAC 847 system (Oxford). The quantity and percentage of slow wave sleep were consistent with those of visual inspection when the minimum delta amplitude was adjusted at ± 25～31μV from the baseline. The gain amplitude of mandibular electromyogram dramatically affected recognition of REM stage by the computer. Percentage of REM sleep was almost same as that of visual inspection when the gain potentiometer was adjusted to read 50μV peak-to-peak calibration signal as 1.0V, while REM stage was not recognized when the gain potentiometer was adjusted to 1.5V. In contrast, the gain amplitude of electropthalmogram did not affect the recognition of REM stage