REPRODUCIBLE EFFECTIVE CONNECTIVITY AND ITS RELATIONSHIP TO COGNITION: A SIMULTANEOUS RESTING-STATE PET/MRI INVESTIGATION

Abstract Background Resting-state functional magnetic resonance imaging (MRI) has been widely applied in neuropsychiatric research. However, the functional connectivity it reveals among brain regions could not convey the direction of signaling flow. Although several statistical methods have been developed to infer effective connectivity (EC) for uncovering signaling direction, their utility is constrained by the inherent limitations of relying solely on MRI data. A novel approach, known as metabolic connectivity mapping (MCM), simultaneously acquires positron emission tomography (PET) and MRI data to assess EC. While the EC derived through MCM has been validated, its reproducibility and relationship to cognition remain uncertain. Aims & Objectives This study aims to investigate the reproducibility of EC determined by MCM and to explore its functional implications concerning cognition. Method We enrolled twenty healthy control participants and acquired resting-state functional MRI and FDG-PET data simultaneously using an integrated scanner at baseline. Eighteen participants underwent a follow- up imaging session one month later. Participants underwent a cognitive assessment probing attention, memory, and executive function at both time points. Group independent component analysis identified four target brain regions: the left middle frontal gyrus, left insula, precuneus, and left angular gyrus. We employed MCM to evaluate the EC between these brain regions. Results Among the twelve pairs of regions assessed for EC, significant EC was observed in seven region pairs at baseline and four region pairs in the follow-up. However, only the bidirectional EC between the precuneus and the left angular gyrus stayed consistent between baseline and follow-up results. Furthermore, the stronger EC from the left angular gyrus to the precuneus compared to the reverse direction was also reproducible. Notably, the EC from the left angular gyrus to the precuneus positively associated with memory function at both baseline (Standardized β = 0.56, p = 0.04) and follow-up (Standardized β = 0.53, p = 0.04) after controlling age and education level. Discussion & Conclusion This study highlights the potential of MCM to explore EC between brain regions. Nevertheless, regional variations in the reproducibility of MCM-derived EC may exist. A comprehensive reproducibility characterization is essential before applying this methodology in future research.