Fig 1.
Experimental paradigm and example of performance from two participants.
(A) Participants performed the control block, consisting of a continuous 1-back categorisation task (left), followed by the performance of the PM block (ongoing task plus the time-based PM, right). Red arrows indicate clock-resets and blue arrows show clock-checks. A clock-reset-trial is highlighted in grey. (B) Example of a high-performance participant. Blue bars represent clock-checks and red bars represent clock-resets, the frequency of clock-checks increases towards the 4-minute target time (right-hand graph). (C) Example of a low-performance participant. Clock-checks were mainly performed at the beginning of the clock-reset-trial (right-hand graph). Note that the length of clock-reset trials was not fixed but depended on each participant’s performance.
Fig 2.
Pipeline for pre-processing and data analysis.
Single subject dataset processing was performed on the continuous data (left of the diagram). Two segmented datasets were created from each continuous dataset: blue pipeline aimed at examining brain sources and event related changes locked to the ongoing task stimulus; pink pipeline aimed at examining clock-checks events. The right of the diagram shows group level processing. Measure Projection Analysis was used to find relevant brain areas and ICs statistically similar across all participants during performance of the ongoing task and the time-based PM task. Four different clustering results were obtained. Event-related analysis was performed on the brain activity associated to the brain domain. ERP: Event-Related Potentials, ERSP: Event-Related Spectral Perturbations.
Fig 3.
Clock reset time and frequency of time-checks.
(A) Clock reset accuracy: Participants were instructed to reset a clock every 4 minutes. Boxplots for each participant in the study show the median and dispersion of clock-reset times. Each circle represents individual clock-resets. The vertical line separates the subjects into two groups. Participants within each group are sorted from smaller to greater interquartile range. (B) Frequency of clock-checks within each clock-reset-trial. High-performers (left) increased number of clock-checks towards the 4-minute target time. Grey lines indicate significant differences, * p < 0.01, ** p < 0.001.
Table 1.
Clock-reset accuracy per group.
Table 2.
Performance in the ongoing task with (PM) and without (control) a prospective memory task embedded.
Fig 4.
Measure Projection Analysis results.
Brain domains revealed by ongoing task events and time-check events (columns), for the two activity measures used in the analysis (rows), ERP and ERSP. Brodmann Areas (BA) are indicated for each result, for details of probability and less probable areas see Table 3. The coloured brain regions represent locations with significant convergence (p<0.01) grouped using a maximum correlation value of 0.8. Note that BA 24 (red domain), here referred to as the MPA-ACC brain domain, is common to all four MPA results.
Table 3.
Anatomical areas associated to each domain for ongoing task and time-check events.
Fig 5.
ERP and ERSP locked to ongoing task events, BA 24.
(A) The left-hand panel shows the Task Condition effect for ERPs (Event types pooled together). The right-hand panel shows the Ongoing Trials effect (Task Condition pooled together). F-values are shown at the bottom of each panel, the red number indicates the threshold F-value after FDR correction at alpha level of 0.05. (B) Task Condition and Ongoing Trials effects are shown in the top and bottom rows respectively. The box outlines indicate alpha (9–15 Hz) and theta frequencies (5–8 Hz). The most right column shows F-values for the significant difference after FDR correction at alpha level of 0.05. Threshold F-Value for each ANOVA result is indicated in the colour bar to the right of the panel. FDR: False Discovery Rate correction.
Fig 6.
ERSP activity of the anterior cingulate cortex for high- and low-performance groups.
Left-hand and middle panels show ERSP to ongoing task events for low- and high-performance groups respectively. Right-hand panel shows a significant group effect (Task conditions pooled together), no significant interaction was found. The box outlines indicate alpha (9–15 Hz) and theta frequencies (5–8 Hz). The most right panel shows F-values for the significant difference after FDR correction at alpha level of 0.05. Threshold F-Value is indicated in the colour bar. FDR: False Discovery Rate correction.
Fig 7.
ERP and ERSP locked to time-checks, BA 24.(A) ERP shows a negative deflection starting about 400ms before time-check. (B) ERSP shows increased alpha/beta desynchronisation starting right after time-check. Time 0 corresponds to the button press, when participants decided to check the clock.
Fig 8.
Power trial-to-trial image for clock-check events.
Trial-by-trial time courses of alpha power. The power of the signal was calculated based on the peak frequency (indicated at the top of each plot) within the frequency band window 9–15 Hz. Time 0 corresponds to button press to check the clock. Trials are sorted by time along the y-axis (black diagonal line, time is scaled to the figure): low numbers in the y-axis corresponds to trials that occurred early within the clock-reset-trial, whereas trials towards the top of the y-axis occurred closer to the 4-minute target time. The left-hand panel shows all participants. Middle and right-hand panels show high- and low-performance group respectively. Time course of the frequency power is shown at the bottom of each panel, when it falls outside the confidence intervals (blue shading), the decrease in alpha power is statistically significant. Trials are smoothed with a 10-trial moving-average. Note that the number of trials in the y-axis is different for each panel.
Table 4.
Summary of specific PM interference task effects across three different PM tasks conditions: Perceptual event-based PM task, conceptual event-based PM task and time-based PM task.