| 摘要: | 睡眠遲惰（sleep inertia，SI）是意識由睡到醒的轉換歷程。晚近國內外功能性磁振造影研究團隊嘗試透過心理動作警覺任務（psychomotor vigilance task，PVT）與靜息態功能性磁振影像（resting state fMRI）了解甫睡醒時的血流動力學反應（hemodynamic responses，HRF）。然而，囿於神經血管耦合（neurovascular coupling）的假設前提，我們了解到睡眠遲惰期間的功能性磁振造影觀察有兩種可能來源：神經元活化或腦血流變化。為了進一步釐清此現象之神經生理基礎，本研究採取閉氣作業以盡可能在排除神經活化的參與下，量測腦血管反應（cerebrovascular reactivity，CVR），以期進一步驗證早前在睡眠遲惰期間，所觀察到的功能性磁振造影結果，所可能代表的生理訊號起源。有鑑於此，透過功能性磁振造影，我們假定睡眠遲惰期間所觀察到的腦區活化的改變，是源自神經基礎。而此改變，並不會受由腦血管反應所捕捉到的血流供給狀況所影響。
實驗設計部份，此夜間睡眠實驗採重複量數設計，利用腦電圖結合功能性磁振造影的同步記錄技術，招募了35位受試者（年齡：24.9 ± 4.1， 7位男性）。實驗期間，主要以夜間睡眠為界切分為四段探討：睡眠前（pre-sleep）及三組睡眠後（A1、A2、A3）。每段約15分鐘，受試者須分別完成警覺任務與腦血管反應量測。在兩種作業中，我們各別安排72個刺激嘗試次（stimulus trial）與重複量測的15秒區塊（block）。之後，根據7個皮質區和3個皮質下的區域取時間序列，挑選對應腦區的警覺任務β權重圖及腦血管反應的動態時間參數。弗理曼二因子等級變異數分析（Friedman test）在A2時警覺任務的反應時間標準差上顯示，和睡前相比，行為表現有弱化趨勢。在10個感興趣腦區的抽值，腦血管反應的時間參數，包含上升時間（ascending time）、替代下降時間（alternative descending time）及半高全寬（FWHM）在視丘（thalamus）上有顯著變化（所有p值在奈梅尼多重比較測定下小於0.05）。爾後，我們試圖計算分段間警覺任務的行為指標和其視丘活化的β權重、視丘β權重和腦血管反應的動態特徵，以及警覺任務的行為指標和腦血管反應特徵的斯皮爾曼等級相關係數（Spearman's rank corre-lation coefficient）。主要的相關性在A2期間的警覺任務的反應時間標準差和腦血管反應的動態特徵間體現（所有未校正的p值均小於0.05）。
警覺任務的行為表現和丘腦的腦血管反應間的相關性，顯示丘腦也許會驅動在警覺任務上所呈現的睡眠遲惰效果。另一方面，在部份皮質區域所觀察到的不同實驗段有顯著變化的血流動力學反應，可能會影響剛睡醒時神經活化的漸進恢復。以上結果潛在指出，在丘腦的血流動力學反應在睡眠遲惰期間可能和皮質區域有所互動，進而影響警醒狀態的恢復。
Sleep inertia (SI) is a transition period upon awakening. Recent fMRI studies ex-plored neural substrate upon SI using psychomotor vigilance task (PVT) and resting state, showing dynamic hemodynamic responses (HRF) of brain activity on awakening. However, due to the presumption of neurovascular coupling, fMRI findings in SI may originate from either the neural or hemodynamic basis. To verify the physiological sub-stratum of SI, we applied breath-hold task to measure cerebrovascular reactivity (CVR) without neural engagements for exploring the physiological origin on previous fMRI findings. We hypothesize that the changing brain activity is originated from the neural basis with unchanged blood supply in SI period, reflected by CVR.
We recruited 35 participants (age: 24.9±4.1, 7 males) for the repeated-measure de-sign in nocturnal sleep experiment using simultaneous EEG-fMRI recordings. In mid-night, four sessions were probed: pre-sleep and 3 post-awakening stages (A1, A2, A3, 15-min apart), and the participants needed to perform both PVT and CVR in each ses-sion. We instructed participants to perform 72 PVT trials and repeated 15-sec breath-hold task for CVR. Next, we extracted time courses within 7 cortical and 3 subcortical regions to extract including PVT β maps and temporal dynamics of CVR. Friedman test showed weaker behavioral performance in PVT SD of RT in A2 relative to Pre. Among 10 prescribed ROIs, thalamus chiefly showed significant CVR changes in ascending time, alternative descending time and FWHM (All p < .05 with Nemenyi multiple com-parison test). Subsequently, we conducted Spearman’s correlation between behavioral indexes and thalamus β, thalamus β estimates and CVR dynamics, and finally behav-ioral indexes and CVR dynamics across sessions. The focal correlations were found between PVT SD of RT and thalamic CVR dynamics in A2 (all uncorrected p < 0.05).
The correlation between PVT performance and thalamic CVR revealed that HRF changes in thalamus might drive the SI effect in response to PVT. On the other hand, the changed HRF in partial cortical regions could influence the progressive recovery of neural activity upon awakening. These results implied that hemodynamic responses in thalamus may interact with cortical activities to determine alert performance during SI. |
