Spatial dependency between task positive and task negative networks

Abstract

Functional neuroimaging reveals both relative increases (task-positive) and decreases (task-negative) in neural activation with many tasks. There are strong spatial similarities between many frequently reported task-negative brain networks, which are often termed the default mode network. The default mode network is typically assumed to be a spatially-fixed network; however, when defined by task-induced deactivation, its spatial distribution it varies depending on what specific task is being performed. Many studies have revealed a strong temporal relationship between task-positive and task-negative networks that are important for efficient cognitive functioning and here. Here, using data from four different cognitive tasks taken from two independent datasets, we test the hypothesis that there is also a fundamental spatial relationship between them. Specifically, it is hypothesized that the distance between task positive and negative-voxels is preserved despite different spatial patterns of activation and deactivation being evoked by different cognitive tasks. Here, we show that there is lower variability in the distance between task-positive and task-negative voxels across four different sensory, motor and cognitive tasks than would be expected by chance - implying that deactivation patterns are spatially dependent on activation patterns (and vice versa) and that both are modulated by specific task demands. We propose that this spatial relationship may be the macroscopic analogue of microscopic neuronal organization reported in sensory cortical systems, and we speculate why this spatial organization may be important for efficient sensorimotor and cognitive functioning.