Abstract
The default mode network (DMN) has recently attracted widespread interest. Previous studies have found that task-related processing can induce deactivation and changes in the functional connectivity of this network. However, it remains unclear how tasks modulate the underlying effective connectivity within the DMN. Using recent advances in dynamic causal modeling (DCM), we investigated the modulatory effect of (gender judgment) task performance on directed connectivity within the DMN. Sixteen healthy subjects were scanned twice: at rest and while performing a gender judgment task. Group independent component analysis was used to identify independent spatial components. Four subject-specific regions of interest (ROIs) were defined according to the ensuing default mode component: the posterior cingulate cortex, the left lateral parietal cortex, the right lateral parietal cortex, and the medial prefrontal cortex. Effective connectivity among these regions was then characterized with stochastic DCM, revealing enhanced (extrinsic) between region connectivity within the DMN during task sessions - and a universal decrease in (intrinsic) self-inhibition - relative to resting sessions. These results suggest a distributed but systematic modulatory effect of cognitive and attentional set on the effective connectivity subtending the DMN: an effect that increases its sensitivity to inputs and may optimize distributed processing during task performance.
Introduction
The Default Mode Network (DMN)—comprising the posterior cingulate cortex (PCC), medial prefrontal cortex (mPFC), and bilateral parietal cortices—is one of the most discussed systems in neuroscience. It is highly active when we are at rest but consistently shows decreased activity (deactivation) when we engage in focused, goal-oriented tasks.
The "puzzle" is why a network so metabolically expensive at rest would simply "shut down" during cognitive effort. While previous studies showed that the correlation between these regions changes during tasks (functional connectivity), they couldn't determine the direction or causality of those influences. This study seeks to uncover the "effective connectivity"—the directed causal influence one region exerts over another—to understand how the DMN's internal architecture is actually modulated by task demands.
Methods and Data
The researchers scanned 16 healthy, right-handed subjects (mean age 20.4) using fMRI under two conditions: a 5-minute resting state and a 5-minute "gender judgment" task (identifying the gender of faces).
To analyze the data, they combined two sophisticated approaches:
- Group Independent Component Analysis (GICA): This was used to statistically separate the fMRI data and identify the spatial map of the DMN for each participant.
- Stochastic Dynamic Causal Modeling (DCM): This is the study's methodological centerpiece. Unlike standard DCM, which requires a known external trigger, stochastic DCM accounts for random fluctuations in brain activity. This allowed the team to model 16 directed connections (both between-region and self-inhibitory) within the DMN.
Key Insights: Deactivation is Not Disconnection
The most striking insight is that the DMN's "deactivation" is not a sign of the network turning off. Instead, the network becomes more internally communicative.
The Excitability Shift
The researchers observed a "universal decrease" in intrinsic self-inhibition within each DMN region during the task. In engineering terms, the "gain" of the system was turned up, making the regions more responsive to each other. By increasing extrinsic (between-region) connectivity and lowering self-inhibition, the DMN moves toward a "critical regime." This state makes the network highly sensitive to sensory inputs or performance monitoring, potentially helping the brain switch back to a "rest" or "monitor" state more efficiently.
Key Results
- Widespread Increases: 13 out of the 16 modeled connections within the DMN showed a significant increase in strength during the task compared to rest.
- Reduced Inhibition: Every region studied showed a decrease in its own self-inhibition parameter during the task.
- Specific Pathways: During the task, the network developed significant bidirectional connections between the parietal regions and the mPFC, as well as a positive connection from the PCC to the mPFC—connections that were not significant during the resting state.
- Causal Correlation: The study found that the strength of these causal connections correlated with the actual BOLD signal changes. For example, the stronger the connection from the mPFC to the left parietal cortex, the greater the observed deactivation in that region.
Limitations and Future Directions
While these findings provide a strong proof-of-concept, the authors acknowledge several areas for further exploration. With only 16 subjects, the results would benefit from larger cohorts to ensure generalizability. Additionally, because the study used a specific gender judgment task, it remains unclear if these connectivity changes are a universal feature of all cognitive tasks or if different types of effort (e.g., math vs. social cognition) modulate the DMN differently.
A natural next step is applying this stochastic DCM approach to clinical populations where the DMN is known to be altered, such as in Alzheimer’s disease or schizophrenia, to see if their effective connectivity fails to reconfigure appropriately during cognitive demands.
Summary updated