OBJECTIVETo establish a brain functional network of the whole brain based on glucose metabolism, and to evaluate the cost and efficiency of the functional network.

METHODS18F-FDG PET of 148 healthy volunteers (30-59 years, n = 148) was performed in a resting state. Images were registered to atlas by using the statistical parametric mapping (SPM) software. The functional connectivity between 90 cortical and sub-cortical regions was estimated by correlation analysis.

RESULTSGlucose metabolism brain functional networks had global efficiency greater than the lattice but less than the random graph, and local efficiency greater than random but less than lattice. This characteristically small-world behavior of the brain network was most consistently seen for low-cost to medium-cost networks. The small world regime was Cost-[0.0512, 0.5406]. The cost efficiency of the networks typically had a maximum positive value when the cost was 0.23.

CONCLUSIONSGlucose metabolism brain functional networks have economical small-world properties. It is feasible to analyze the functional characteristics of human brain by study 18F-FDG PET images. This paper provides a new method for studies on brain function.