Alpha and theta peak frequency track on- and off-thoughts

Figures

• 

  Fig. 1. Experiment process and analysis schema. Experiment process: Each participant performed 1800 Sustained Attention to Response Task (SART) trials. In each trial, three types of stimuli were presented randomly for 100 ms after 200 ms fixation period: standard stimulus (upper case English letters), target stimulus (lower case English letters) and novel stimulus (letters from minority languages). Participants were requested to respond to the target stimuli by pressing the F key on the keyboard during the 900 ms blank window following stimulus (post-stimulus period). After each 15th target stimulus two thought probes were shown, participants were asked to answer the probe questions based on their types of thoughts (on- vs off-task and single vs multiple contents).

• 

  Fig. 2. Behavioral analysis schema and results. a Analysis schema: Data analysis was trial-based across subjects. The accuracy, response time (RT), and probe answers were taken into behavioral analysis. All trials were divided into different conditions according to probe answers: off-thought vs on-thought based on the first probe, multiple-contents vs single-content based on the second probe, off-multiple vs on-single based on the combination of the first as well as second probe. GLMM and LMM were applied on accuracy and RT between conditions respectively. Chi square analysis was applied between two probes. b Results of GLMM and LMM on the accuracy and RT between off-multiple and on-single. There were no significant differences between off-multiple and on-single on accuracy or RT. The error bar means SD; ns: no significance.

• 

  Fig. 3. EEG analysis schema. EEG signals were first transferred to peak frequency sliding (FS) and frequency power (FP), then the average of pre-stimulus period (−200–100 ms) was subtracted from post-stimulus period (100–600 ms) to obtain the relative values. Alpha (8–13 Hz) and theta (5–8 Hz) post-stimulus relative values were compared between conditions, i.e., the different thought types, using LMM.

• 

  Fig. 4. Frequency sliding (FS) and frequency power (FP) during different thought types. a The difference map of FS post-stimulus’ relative values. The difference map is calculated by subtracting on-single values from off-multiple values. According to visual inspection showing the strongest changes of FS, Fz was chosen for alpha FS and FP, CPz was chosen for theta FS and FP. b The time series of FS and FP relative values (subtracting the average of −200–0 ms from all time points) under each condition (off-multiple and on-single) with the results of LMM of post-stimulus relative values between off-multiple and on-single. Significant differences were found between off-multiple and on-single thoughts in both alpha and theta FS but not in FP. *p < 0.05; **p < 0.01; ns: no significance.

• 

  Fig. 5. Results of FS and FP dynamic time warping (DTW). a The result of FS and FP DTW. The distances between off-task and multiple-contents, on-task and single-content are the smallest on both alpha and theta FS. The lines mean there’re significant differences between groups. b The schematic graph of the DTW result.

• 

  Fig. 6. The joint distribution of FS alpha and theta and joint entropy (JE). a The joint distribution of alpha/theta FS (and FP). Off-multiple and on-single’s distribution could be discriminated in FS with overlaps. b The joint entropy (JE) results in FS. JE values for off-multiple thoughts are significantly higher than those for on-single thoughts. The error bar means SD. ****p < 0.0001.

• 

  Fig. 7. Schematic illustration of the proposed mechanisms mediating the relationship between neuronal input, alpha/theta frequency sliding and thought types. a - On-single thoughts. Here during external tasks, the single external input (yellow) dominates, activates external-neurons on the population level (lower level of graph), increases alpha peak frequency, decreases theta peak frequency (middle level), and leads to the cognition of on-single thoughts (upper level). b - Off-multiple thoughts. Here, multiple internal inputs dominate during external tasks (lower level) (blue), increase theta peak frequency, decrease alpha peak frequency (middle level), and are psychologically manifest as off-multiple thoughts (upper level).