Dissecting Neural Responses to Temporal Prediction, Attention, and Memory: Effects of Reward Learning and Interoception on Time Perception

Abstract

Temporal prediction (TP) is needed to anticipate future events and is essential for survival. Our sense of time is modulated by emotional and interoceptive (corporal) states that are hypothesized to rely on a dopamine (DA)-modulated "internal clock" in the basal ganglia. However, the neurobiological substrates for TP in the human brain have not been identified. We tested the hypothesis that TP involves DA striato-cortical pathways, and that accurate responses are reinforcing in themselves and activate the nucleus accumbens (NAc). Functional magnetic resonance imaging revealed the involvement of the NAc and anterior insula in the temporal precision of the responses, and of the ventral tegmental area in error processing. Moreover, NAc showed higher activation for successful than for unsuccessful trials, indicating that accurate TP per se is rewarding. Inasmuch as activation of the NAc is associated with drug-induced addictive behaviors, its activation by accurate TP could help explain why video games that rely on TP can trigger compulsive behaviors.

Keywords: dopamine; fMRI; reward; timekeeping.

Published by Oxford University Press 2014. This work is written by (a) US Government employee(s) and is in the public domain in the US.

Figures

Figure 1.

(A) Time course of the TOP task, highlighting the onset and durations of the visual cues (brown), TD learning, response window (green), and the presentation of outcome (pink) within one interstimulus interval (ISI = 11–13 s). (B) Spatiotemporal distribution of events, targets, responses, and outcomes for the TOP task within the ISI, highlighting the nature and timing of the cues as well as the TD learning. Dashed arrows indicate the sequential advent of circles in the screen; dashed arrows are not shown in the screen during the task. (C) TD learning module: After the first circle is displayed on one of the corners of the screen, the 3 remaining circles are sequentially displayed at fixed intervals at the corners of the screen, as indicated by the cue in B. The participants predict the timing of each circle [W(t), FWHM Gaussian solid color curves], which might not match perfectly the correct timing of the circles [V(t), FWHM Gaussian black solid curve centered at t1, t2, and t3]. The error function, δ(t), measures the time difference between the predicted events and the correct events (dotted color curves). The MRI signal, assumed to be proportional to the time integral of δ(t), is expected to be higher for unsuccessful (miss) than for successful (hit) prediction trials.

Figure 2.

Column scatter graph showing the average behavioral responses across trials for each participant (circles) during the SM (blue), TP (green), WM (red), and SA (pink) fMRI tasks. RT was significantly different for all tasks (P < 5 × 10−6; paired t-test). Performance accuracy was significantly different for all tasks (P < 5 × 10−6), except for the WM and SA which did not show significant differences in accuracy (P > 0.08).

Figure 3.

Statistical significance of brain activation differences superimposed on axial and sagittal views of the human brain (radiological convention) showing the increased activation during the TOP task, compared against the SM (A), WM (B), and SA (C) tasks as well against the co-activation caused by the SM, WM, and SA tasks (D); one-way within-subject ANOVA. (E) Column scatter graph showing the average BOLD responses elicited by each of the fMRI task within 27 voxels cubic ROI centered at the locations of the medial thalamus, right anterior insula (bilateral), and right NAc (see MNI coordinates of the ROIs in Table 1).

Figure 4.

Statistical significance of brain activation evoked by the TD learning and outcome epochs of the TOP task and their difference (TD learning > outcome) superimposed on axial views of the human brain (radiological convention).

Figure 5.

(A) Association across subjects between RT and activation responses in the cerebellum and right anterior insula during the TOP task (linear regression; df: 34). (B) Association between RT and activation responses in the right anterior insula and NAc (linear regression; df: 106) across subjects and tasks. Statistical significance (red–yellow color maps) is superimposed on axial views of the human brain (radiological convention).

Figure 6.

(A) Column scatter graph showing the average RT for successful (hits; green circles) and unsuccessful (misses; red circles) trials during the TOP task; (B) statistical significance (t-score) of differential activation responses for successful (hit) and unsuccessful (miss) trials, independently for TD learning and outcome epochs, superimposed on coronal and axial views of the human brain in the radiological convention (within-subject ANOVA); (C) column scatter graph showing the average BOLD responses in NAc elicited by TD learning and outcome, independently for hit and miss trials; (D) average BOLD responses in the cerebellum (CB) and VTA elicited by TD learning, independently for hit and miss trials (27-voxel cubic ROI centered at x, y, z = 0, −13, −12 mm).

Figure 7.

Linear association across subjects between RT and TD learning activation responses in VTA, and cerebellum independently for hit and miss trials (df: 34).