| 摘要: | 生物體能夠學習在環境中較有可能發生，或是對該生物體有利的事件，並反應在接下來的行為表現中。例如在搜尋作業中，受試者就可以學習到目標物較有可能出現的位置。動物可以根據現有的信息，調整他們的注意力資源分配特定事件。舉例而言：一些注意力投注而造成的效果，像是逆向眼動耗損（antisaccade cost），就會受機率事件影響而改變。然而，生物體對於某些事件的期望是如何去影響他們的行為，和其潛在的神經機制仍不清楚。期望值可以藉由事件發生的機率和酬賞的大小而量化的操弄。本研究將對機率與酬賞分別進行操弄。研究的第一部分中，我們使用三個跨顱直流電刺激（tDCS）實驗來探討空間機率所造成的效果和其相關的神經機制。第二部份的實驗則探討酬賞的大小對行為造成的影響。
在實驗一和實驗二中，我們分別施打正極tDCS在額葉眼動區（rFEF）以及額葉輔助眼動區（SEF）。而在實驗3中，則將負極tDCS施打於額葉輔助眼動區。在tDCS的刺激後，受試者須要完成含有空間機率操弄的眼動作業。眼動作業中包含了正向眼動（prosaccade）和逆向眼動（antisaccade）兩種眼動型態。實驗一和實驗二的結果顯示，當正極的tDCS施打在額葉眼動區會使正向眼動的反應眼動時間（saccade latencies）減短。此外，相對於高機率位置，正極tDCS在低機率位置對於反應眼動時間有較大的促進效果。這兩種促進的效果都只發生於對額葉眼動區進行刺激而非額葉輔助眼動區。在實驗3中，負極tDCS施打在額葉眼動區使逆向眼動的反應眼動時間減短。這三個實驗顯示額葉眼動區相對於額葉輔助眼動區，在起始眼動上的重要性，並且確認了此區域在處理空間機率所扮演的角色。
在第二部分的研究中，我們在實驗四以及實驗五中操弄獎勵的大小，並讓受試者分組進行正向眼動以及負向眼動作業。在這兩種眼動類型中可以分別探討注意力和運動準備投注到相同或是相反的位置時，酬賞在此兩種情況下所造成的影響。實驗結果證實了在視覺選擇作業下的酬賞對反應眼動時間所造成的促進效果，另外，也提供了酬賞效果在眼動作業中可能造成促進效果的階段。綜上所述，結果顯示兩個在期望值當中扮演重要上對下控制（top-down control）的因素，機率和酬賞，在眼動準備的注意力分配上都為重要的成分。當下累積之期望值所造成的增益除了發生在運動準備的階段外，也可增強在視覺注意力階段的調節。
Animals are capable of learning occurrences that may happen in the environment, and the knowledge of their learning is reflected in their following behavior. According to the prior information, animals can adjust their attention allocations to specific events. For example, performance in searching task and antisaccade cost can be affected by probability events. However, how expectation modulates participants’ behavior and their underlying neural mechanisms are still unclear. Expectation can be measured as the probability for an action and the magnitude of the reward. The present study will manipulate probability and reward size separately. In the first part of this study, three tDCS experiments were utilized to establish the causal links between brain regions and location probability effect. The contribution of reward effect is examined in the behavior experiment in the second part.
In experiment 1 and 2, offline anodal (excitatory) tDCS was applied with the active electrode placed over the rFEF or SEF. In experiment 3, offline cathodal (inhibitory) tDCS was applied over rFEF. Directly after tDCS stimulation, participants were asked to complete a pro- or anti- saccade task which manipulated location probability. The results in experiment 1 and 2 indicated that saccade latencies for prosaccades were significantly shorter when anodal tDCS was applied to the rFEF but not SEF. Additionally, when applying anodal tDCS in FEF produced a larger effect for low probability locations than high probability locations led a major interaction between tDCS condition and location probability. In experiment 3, cathodal tDCS led faster saccade latencies in antisaccades. The results demonstrate that reduction on saccade latencies induced by both anodal and cathodal tDCS over the rFEF, which confirms the role of rFEF in saccade initiation. The findings also dissociate the critical roles of the rFEF and SEF in the effect of location probability and confirm the importance of rFEF in processing location probability information. The current tDCS experiments suggest that the FEF plays a critical role in modulating the location probability effects and the saccade latencies.
In the second part of the study, we investigated other factor that is essential to expectation. In experiment 4 and 5, reward size was manipulated in the prosaccade task and antisaccade task. The separation of two saccade types could exam the effect of reward when the spatial attention and motor preparation is compatible or not. The results confirm the reward effect in a visual selection task. Moreover, we provided the potential stages that top-down control might influence in saccade latencies.
In summary, the results show that two top-down factors are related to expectation, probability and reward, can account for attentional allocation of saccadic preparation. The prior expectation is not only an important factor for motor preparation, but also critical for the flexibility of visual attention. |
