The Effect of Seed Location on Functional Connectivity: Evidence from an Image-Based Meta-Analysis

Abstract Default mode network (DMN) is the most involved network in the study of brain development and brain diseases. Resting-state functional connectivity (rs-FC) is the most used method to study DMN, but different studies are inconsistent in the selection of seed. To evaluate the effect of different seed selection on rs-FC, we conducted an image-based meta-analysis (IBMA). We identified 59 coordinates of seed regions of interest (ROIs) within the default mode network (DMN) from 11 studies (retrieved from Web of Science and Pubmed) to calculate the functional connectivity; then, the uncorrected t maps were obtained from the statistical analyses. The IBMA was performed with the t maps. We demonstrate that the overlap of meta-analytic maps across different seeds’ ROIs within DMN is relatively low, which cautions us to be cautious with seeds’ selection. Future studies using the seed-based functional connectivity method should take the reproducibility of different seeds into account. The choice of seed may significantly affect the connectivity results.