Skip Navigation

State Dependence of Functional Connectivity Networks

Author: Gopinath Kaundinya, PhD

Functional connectivity MRI (fcMRI) uses coherent resting state fluctuations in blood oxygenation level dependent (BOLD) functional magnetic resonance imaging (fMRI) data to infer functional brain networks. FcMRI has been employed to study functional connectivity networks in healthy brain and in disease. BOLD fcMRI networks are believed to reflect

  1. Either anatomically constrained spontaneous fluctuations,
  2. Or state-dependent activity independent of anatomic constraints,
  3. Or some combination of both types of brain activity.

In this study, we examined the state-dependence of BOLD fcMRI networks by examining the hypothesis that functional connectivity along specific brain networks will be modulated by the resting state condition used to assess them. In this regard, we made use of the classical duality in cognitive and emotion systems, wherein increased brain activity in one (say cognitive) system results in a corresponding decrease in brain activity in the other (emotion). The striatum has well-known projections to both cognition and emotion systems. The dorsal striatum (DS) is more strongly connected to sensory and attention networks and ventral striatum (VS) is more connected to ventral affective processing networks. VS also acts as an interface between limbic and sensorimotor/cognition systems.

We examined resting state functional connectivity to DS and VS in 13 right–handed subjects under two different fMRI conditions.

  1. Resting Eyes Open (REST): unstructured and self-regulated “neutral” affective condition.
  2. Transcutaneous bilateral median nerve stimulation (TENS): Condition engages sensory and attention systems

The hypotheses were

  1. Dorsal striatum functional connectivity to sensory and attention areas will increase in strength from REST to TENS.
  2. Ventral striatum functional connectivity to affective processing areas will decrease in strength from REST to TENS.
  3. Ventral striatum functional connectivity to sensory and attention areas will also increase in strength from REST to TENS.

Figures 1 and 2 show the results of the study. Dorsal striatum functional connectivity (Figure 2) with sensory and attention areas (primary somatosensory cortex, paracentral lobule, dorsal anterior cingulate and dorsolateral prefrontal cortex) went up from REST to TENS as hypothesized. At the same time ventral striatum functional connectivity (Figure 1) with ventral affective process areas went down from REST to TENS consistent with hypothesis of reciprocal behavior of striatal functional connectivity to cognition and emotion systems. Since VS also has projections thru DS to sensory and attention areas, VS functional connectivity to these areas also increased from REST to TENS. The results add to existing evidence that suggests that resting state fcMRI networks reflect in a large part state-dependent activity.

Figure 1

Figure 1: Ventral striatum (VS) shows significantly (p < 0.02) weaker functional connectivity with ventral affective processing areas (blue) during minimally perceptible continuous transcutaneous bilateral median nerve stimulation (TENS) compared to resting eyes open condition (REST). On the other hand VS exhibits significantly (p < 0.02) stronger functional connectivity with sensory and attention areas (red) during TENS than in REST.

Figure 2

Figure 2: Dorsal striatum shows significantly (p < 0.02) stronger connectivity with sensory and attention areas during TENS compared to REST.

Future work in our lab will focus on apportioning to what extents fcMRI networks reflect state-dependent and anatomically constrained spontaneous fluctuations, by creative use of brain imaging techniques as well as paradigm design. The potential of differential state dependence of functional connectivity networks in normal and diseased populations will also be examined, with the aim of developing appropriate biomarkers based on these investigations.

Project Team

Kaundinya Gopinath (PI), TBA Postdoc