| ObjectiveMind wandering is referred as to drift away from a environment-information processing task toward unrelated thoughts, fantasies, feeling, and other musings. The focus of attention to the task can't be sustained. Mind wandering often led to failures to the task, and the reaction time of error responses was faster. Antrobus (1968) presented that during a simple signal detected task, the rate of occurrence for mind wandering decreased while the probability of rare stimuli was high. This was because executive control and mind wandering completed limited cognitive resources. Weissman et al. (2006) found that attention lapse was characterized by increased activity in widespread regions of right frontal and anterior cingulate cortex, less deactivation of the default mode network. O'Connell et al. (2009) reported that electrophysiological signals predicted errors up to 20s before they occurred. Small wood et al. (2008) found that the P300 amplitude appeared to be smaller for rare stimuli in the "error" relative to "correct" condition by ERP and sustained attention to response task. It implied that there would be a reduction in the depth of the cognitive analysis to the environment information when the mind wanders. The neural mechanism of how to affect the distribution of cognitive resources by the stimulus probability is less investigation.We defined operationally that mind wandering was the error responses to the task. We aimed to investigate the neural mechanism of mind wandering by changing the probability ratio and spatiotemporal pattern analysis of ERP. MethodsTen right-handed subjects (6 female) participated without neurological or mental disorders. Their ages ranged from 23 to 26 (24.40±1.07 years). They were all provided written informed consent.We used the sustained attention to response task and the block design. The stimulus materials consisted of four kinds of Flanker pictures, which were separately composed by five arrows. The conflicting symbol string of both ">><>>" and "<<><<" was frequent stimuli that were asked to be have a response according to the direction of the middle arrow. In order to increase the rate of false response, subjects were asked to response according to the opposite direction of the middle arrow to the consistent symbol string of both "<<<<<" and ">>>>>" (the rare stimuli). Every picture was lasted for 300 ms with the interval between stimuli of 700 ms. The probability structure of 0.1/0.9 consists of 180 rare stimuli and 1620 frequent stimuli. The other two groups are characteristic with the probability structure of 0.2/0.8 and 0.3/0.7 respectively. The order of blocks was balanced between subjects.EEG was recorded from 19 channels of international 10-20 system. Event-related EEG epochs from -100 ms before to 900 ms after the onset of the frequent stimulus were averaged offline. The ERP waveforms averaging from the 1-5 EEG epochs of the frequent stimuli was classified into wandering (response to the rare stimulus correctly) and no-wandering (response to the rare stimulus erroneously). The missing rate for rare stimuli, the sensitivity d'and the reaction time for mind wandering and no mind wandering were recorded.For statistical analysis, ERPs elicited by the frequent stimuli and the reaction time were analyzed with two-way ANOVA of repeated-measurement (wandering:yes, no)x(SART probability ratio:0.1/0.9,0.2/0.8,0.3/0.7). The missing rates were assessed by a one-factor repeated-measure (SART probability ratio:0.1/0.9,0.2/0.8, 0.3/0.7) ANOVA with SPSS 13.0. The simple effect analysis corresponding to the SART probability ratio and the wandering conditions separately and a paired, two-sided t-test as post hoc for each condition were further performed. The significance level was 0.05.The ERP data of each channel at every moment was submitted to the same statistical comparison. Conservative lower-bound epsilon (0.5) was applied to adjust the degree of freedom of the task factor since its simplicity. The resultant multichannel time series of both F-value an t-value were used to generate spatiotemporal patterns by an interpolation method of generalized cortical imaging technique, which was referred as statistical parametric mapping of F-value or t-value of ERP, in the abbreviation form as SPM(F) and SPM(t) separately.ResultsBehavior performanceThe RT interaction effect is not significant, but the mind wandering effect (F(1, 9)=38.108,P=0.000) and the probability effect (F(2,18)=14.083, P=0.000) are significant. The followed LSD analysis confirmed significant differences for the 0.1/0.9 probability vs. the other two probabilities. The probability effect for missing rate is significant (F(2, 18)=4.924, P=0.020), the missing rate of the 0.1/0.9 probability is higher than the other two probabilities. The probability effect of sensitivity d'is not significant (F(2, 18)=0.825,P=0.454).Spatiotemporal patterns of SPM (F)Significant interaction effects of mind wandering and probability occurred in bilateral parietal areas (500-550 ms,150-200 ms and 500-600 ms) and right temporal-occipital cortex (200-250 ms). Significant probability effects appeared in left prefrontal and right parietal areas (300-350 ms) and bilateral frontal cortex (800-900 ms). Significant mind wandering effects also occurred during four periods separately: bilateral occipital cortex (150-300 ms), bilateral frontal pole (250-300 ms), bilateral frontal-temporal-parietal cortex (400-600 ms), and left prefrontal and right frontal cortex (800-900 ms).Spatiotemporal patterns of SPM(t)Pair wise comparisons between mind wandering and no mind wandering for the three probability structure showed that:Significant mind wandering effects occurred in bilateral occipital cortex (250-300 ms), bilateral temporal-parietal- occipital cortex (400-550 ms), and bilateral frontal-central-parietal region (350-550 ms).ConclusionIn line with Smallwood's findings, the RT results were significantly slower under the higher rare stimuli probability condition and speeding of responses was predictive of subsequent errors in the SART. It is argued that the high frequency of targets acts as external support to performance and hence reduces the need for internal attentional allocation to response selection. The probability effect of d'showed that mind wandering mostly affects the perceptual process.The probability effect in left prefrontal cortex and right parietal cortex (300-350 ms) presented that top-down attentional system was affected by the probability. Top-down system can affect mind wandering by distributed cognitive resources, which can be proved by the interaction effects of mind wandering and probability in bilateral parietal cortex (500-550 ms).Smallwood (2008) only reported that the P300 amplitude was reduced during mind wandered. Our study showed that mind wandering occurred at every period of cognitive process:bilateral occipital (150-350 ms) and frontal pole (250-300 ms), thought to be associated with visual process; bilateral frontal-temporal-parietal cortex (400-600 ms), thought to be associated with judgment and determination; left prefrontal-temporal and right frontal cortex (800-900 ms), thought to be associated with monitoring conflict. The wandering effect in bilateral frontal cortex (800-900 ms) the probability effect were independent, which presented that monitoring failure may be an important part to the occurrence of mind wandering.Braboszcz et al. (2011) presented that mind wandering resulted in the reduced level of alertness, which can interpret that mind wandering had a wide influence on the whole cognitive process. During mind wandering, attention drifted away from the task towards inner thoughts, so there was a reduction in the depth of the cognitive analysis about environment and the response was prone to be wrong. On the other hand, mind wandering may be the result of familiarity. The functional failure of the executive attention system influenced by the probability structure of SART may cause the occurrence of mind wandering. Mind wandering may reduce the activation of cognitive processing for environment information and cause speedy response with more faults. How to distinguish automatic process and mind wandering elicited by fatigue will be an important subject in the future. |
PDF Full Text Request