Synchronization of medial temporal lobe and prefrontal rhythms in human decision making

Abstract

Optimal decision making requires that we integrate mnemonic information regarding previous decisions with value signals that entail likely rewards and punishments. The fact that memory and value signals appear to be coded by segregated brain regions, the hippocampus in the case of memory and sectors of prefrontal cortex in the case of value, raises the question as to how they are integrated during human decision making. Using magnetoencephalography to study healthy human participants, we show increased theta oscillations over frontal and temporal sensors during nonspatial decisions based on memories from previous trials. Using source reconstruction we found that the medial temporal lobe (MTL), in a location compatible with the anterior hippocampus, and the anterior cingulate cortex in the medial wall of the frontal lobe are the source of this increased theta power. Moreover, we observed a correlation between theta power in the MTL source and behavioral performance in decision making, supporting a role for MTL theta oscillations in decision-making performance. These MTL theta oscillations were synchronized with several prefrontal sources, including lateral superior frontal gyrus, dorsal anterior cingulate gyrus, and medial frontopolar cortex. There was no relationship between the strength of synchronization and the expected value of choices. Our results indicate a mnemonic guidance of human decision making, beyond anticipation of expected reward, is supported by hippocampal-prefrontal theta synchronization.

Figure 1.

Memory and decision-making task. A, Example of the active choice condition (experimental task). B, Example of the forced choice condition (control task). C, Example timeline of the experiment. D, Probability of making a correct choice across all experimental blocks (mean ± SEM). The x-axis represents the trial index relative to the last error trial before learning occurred (highlighted in red) that was defined using an arbitrary criterion (last error trial before participants performed six or more consecutive correct choices). Participants required on average 8.31 trials to reach this criterion (highlighted in green), and during this period performance was at chance. The first six trials after the criterion did not involve any error trial, because the learning criterion was determined based on the performance of six correct trials (highlighted in blue).

Figure 2.

Time frequency analysis. A, B, The topographical maps of the T values (sensor space) and thresholded (p < 0.05; whole brain FWE) T maps in the time domain that result from contrasting active versus forced choice trials show higher induced theta power in the active choice conditions over frontal and centrotemporal sensors. C, D, Group-averaged time frequency activity (pooled from all channels included within the highlighted areas in A and B) for the active choice, the forced choice, and their difference. E, Anatomical localization of the sources of the differences in theta power between active and forced choice conditions detected a medial prefrontal cortex and an anterior hippocampal cluster (p < 0.05, FWE). The color scale indicates T values.

Figure 3.

Theta power in medial temporal lobe correlates with behavioral performance. In the medial temporal lobe source, the difference in theta power between experimental and control condition during the decision period was negatively correlated with the number of decision errors (R² = 0.421; p = 0.002).

Figure 4.

Quantitative eye movement analysis. Probability of a saccade across the decision period is shown. On average, a higher frequency of saccades was observed in the control than in the experimental condition. The line on the top highlights significant differences between control and experimental conditions on a related samples t test (p < 0.05).

Figure 5.

Theta phase synchronization between the hippocampus and the prefrontal cortex during decision making in the postlearning phase. A, Phase locking index, PLI, of the theta rhythm to the medial temporal lobe theta rhythm was higher in the active choice conditions when compared to the forced choice condition in the lateral superior frontal gyrus (p < 0.05, whole brain FWE). PLI was calculated within a time window spanning 200 to 2000 ms after the start of the decision period. B, Time course of PLI for the active and forced choice conditions for the peak voxel identified in A. For this analysis, the original time window was divided into three 600 ms time windows. C, PLI of the theta rhythm to the medial temporal lobe theta rhythm was higher in the active choice conditions when compared to the forced choice condition in the dorsal anterior cingulate cortex (p < 0.05, whole brain FWE). PLI was calculated within a time window spanning 200 to 2000 ms after the start of the decision period. D, Time course of PLI for the active and forced choice conditions for the peak voxel identified in C. For this analysis, the original time window was divided into three 600 ms time windows. E, PLI of the theta rhythm to the medial temporal lobe theta rhythm was higher in the active choice conditions when compared to the forced choice condition in the medial frontopolar cortex (p < 0.05, whole brain FWE). PLI was calculated within a time window spanning 200 to 2000 ms after the start of the decision period. F, Time course of PLI for the active and forced choice conditions for the peak voxel identified in E. For this analysis, the original time window was divided in three 600 ms time windows.