PT  - JOURNAL ARTICLE
AU  - Murat Demirtaş
AU  - Matthieu Gilson
AU  - John D. Murray
AU  - Dina Popovic
AU  - Eduard Vieta
AU  - Luis Pintor
AU  - Vesna Prčkovska
AU  - Pablo Villoslada
AU  - Gustavo Deco
TI  - Exploring Anti-correlated Resting State BOLD Signals Through Dynamic Functional Connectivity and Whole-brain Computational Modeling
AID  - 10.1101/085274
DP  - 2016 Jan 01
TA  - bioRxiv
PG  - 085274
4099  - http://biorxiv.org/content/early/2016/11/02/085274.short
4100  - http://biorxiv.org/content/early/2016/11/02/085274.full
AB  - Resting-state functional magnetic resonance imaging and diffusion weight imaging became a conventional tool to study brain connectivity in healthy and diseased individuals. However, both techniques provide indirect measures of brain connectivity leading to controversies on their interpretation. Among these controversies, interpretation of anti-correlated functional connections and global average signal is a major challenge for the field. In this paper, we used dynamic functional connectivity to calculate the probability of anti-correlations between brain regions. The brain regions forming task-positive and task-negative networks showed high anti-correlation probabilities. The fluctuations in anti-correlation probabilities were significantly correlated with those in global average signal and functional connectivity. We investigated the mechanisms behind these fluctuations using whole-brain computational modeling approach. We found that the underlying effective connectivity and intrinsic noise reflect the static spatiotemporal patterns, whereas the hemodynamic response function is the key factor defining the fluctuations in functional connectivity and anti-correlations. Furthermore, we illustrated the clinical implications of these findings on a group of bipolar disorder patients suffering a depressive relapse (BPD).