用眼動馬可夫模型與事件相關電位分析思緒漫遊和專注

Mind-wandering revealed by eye movement hidden Markov model and ERPs

陳子翎(Zih-Ling Chen)

指導教授 ： 葉素玲

國立臺灣大學/醫學院/腦與心智科學研究所/碩士(2021年)

https://doi.org/10.6342/NTU202100507

摘要

隨著現代科技和社群媒體的蓬勃發展，人們身處於一個分心時代，越來越難以專注在當前重要的工作上。當我們把注意力轉移到與作業無關的思緒上，這樣的現象又稱為思緒漫遊，它與作業表現之間有顯著的負向關聯，因此探討思緒漫遊是重要的議題。先前的研究已指出專注和思緒漫遊兩種不同注意力狀態下，有不同的眼動行為，然而目前仍不清楚是否能藉由眼動行為區分不同的注意力狀態。本研究因此想探討是否能透過眼動模式，將人們分成傾向專注和傾向思緒漫遊。在此研究中，受試者會執行持續性注意力作業同時也記錄事件相關電位。持續性注意力作業共有40個回合，每個回合有25個嘗試次數，其中包含一個需要停下的目標刺激和24個需要按鍵反應的非目標刺激。每個回結束時會請受試者自評當下的注意力狀態。本研究利用眼動馬可夫模型將受試者區分成兩種不同的眼動模式：集中型和分散型，並且比較不同眼動模式下的行為表現與事件相關電位。研究以目標刺激出現的行為表現為基準，將目標刺激出現之前的10秒作為客觀量測思緒漫遊的時間窗。結果顯示與使用分散型眼動模式的受試者相比，使用集中型眼動模式的受試者有較好的作業表現。同時當受試者錯誤反應目標刺激的當下，量測到P3振幅減少以及N2振幅變大，另外非目標刺激的P3振幅也是減少的。研究也以自評問題的回答為基準，將自評問題出現之前的10秒作為主觀量測思緒漫遊的時間窗。結果顯示使用集中型眼動模式的受試者傾向回答專注的比例較高，然而沒有發現回答專注和思緒漫遊之間的非目標刺激P3振幅的顯著差別。這些結果表明，專注和思緒漫遊之間存在特定的眼動模式，P3振幅的減少表明思緒漫遊時降低對作業的認知處理。 本研究揭示眼動和注意力之間的關聯，強調眼動可以區辨不同的專注狀態，也開啟眼動馬可夫模型於視覺注意力研究的新視野。

關鍵字

思緒漫遊 ； 注意力 ； 眼動模式 ； 眼動馬可夫模型

並列摘要

With the development of modern technology and social media, we live in a distracted era that makes us more difficult than ever before to concentrate on the current goal and thus important tasks at hand. It has become imperative to identify mind-wandering, a phenomenon that people sometimes think about things unrelated to the current task, which can potentially cause considerable negative effects on task performance. Since eye movements have been shown to reveal different characteristic between focused attention and mind-wandering, we examined whether eye movement patterns can categorize different groups of people by how prone they are to mind-wandering. Participants performed the sustained attention to response task (SART) with their eye movements and ERPs recorded. The SART comprised 25 trials (one No-go target and 24 go trials) per block and 40 blocks in total. At the end of each block, participants were asked to subjectively rate their state of attention. By applying the eye movement hidden Markov model (EMHMM) to analyze the eye movement data, we differentiated people with two eye movement patterns: centralized vs. distributed pattern. We analyzed the 10-s pre-target time window, which served as an objective measure of mind-wandering based on their performance on the No-go target. Results showed that participants with a centralized-viewing pattern had higher d’ than those with a distributed-viewing pattern. Also, the incorrect response to the No-go target were associated with decreased P3 and greater N2. P3 amplitudes for go trials was also reduced prior to the incorrect response to No-go target. The 10-s pre-probe time window was also analyzed, which served as a subjective measure of mind-wandering based on their report on the probe. Participants with a centralized-viewing pattern tended to rate themselves as more focused relative to those with a distributed-viewing pattern. However, no difference on P3 for go trials prior to rating the state of focus was found. These results suggest that there are specific eye movement patterns differentiating focused attention and mind-wandering, and P3 amplitude reduction suggests reduced processing of the task during mind wandering. By linking the relationship between eye movements and attention, this study highlights that states of attention can be revealed by eye movements and also provides new insight in utilizing EMHMM to study visual attention.

並列關鍵字

mind wandering ； attention ； eye movement pattern ； EMHMM

參考文獻

Aston-Jones, G., Cohen, J. D. (2005). An integrative theory of locus coeruleus-norepinephrine function: adaptive gain and optimal performance. Annu. Rev. Neurosci., 28, 403-450.

Bates, D., Sarkar, D., Bates, M. D., Matrix, L. (2007). The lme4 package.

Beatty, J., Lucero-Wagoner, B. (2000). The pupillary system. Handbook of psychophysiology, 2(142-162).

Braboszcz, C., Delorme, A. (2011). Lost in thoughts: neural markers of low alertness during mind wandering. NeuroImage, 54(4), 3040-3047.

Ceh, S. M., Annerer‐Walcher, S., Körner, C., Rominger, C., Kober, S. E., Fink, A., Benedek, M. (2020). Neurophysiological indicators of internal attention: An electroencephalography–eye‐tracking coregistration study. Brain and behavior, 10(10), e01790.