EEG‐based Analysis Reveals Different Effective Connectivity Patterns between Healthy Controls and MCI Patients in Motion Detection Tasks

Abstract

AbstractBackgroundBrain connectivity has been proposed to identify people at risk for mild cognitive impairment (MCI). Existing work has mainly focused on functional connectivity which reflects the mutual dependence between brain regions. Effective connectivity, which represents the directed connectivity from one brain region to the other, is expected to provide more details on directed information flow in the brain network. The present study investigates how cognitive engagement changes the effective connectivity in the brain.MethodOur research focuses on task‐based EEG (64‐channel), where participants were asked to perform a motion direction discrimination task. The current dataset includes 56 consensus‐diagnosed, community‐dwelling African Americans (ages 60‐90 years, 28 healthy controls (HC) and 28 MCI patients) recruited through the Wayne State Institute of Gerontology and Michigan Alzheimer’s Disease Research Center with subjective cognitive complaints.We evaluated the effective connectivity at different time periods of the motion‐detection task across all the possible EEG region pairs using causalized convergent cross‐mapping (cCCM) of the current source density. For each task trial, the successive time periods being examined included: (I) stimulus onset to Go indication, (II) Go indication to motion‐stop, and (III) the button‐press period. For each subject, we calculated the mean, maximum and minimum of the cCCM vector for each EEG region pair and each time‐period, then selected the region pairs where the cCCM‐based effective connectivity of HC and MCI show a significant difference (p‐value<0.05).ResultOur analysis indicated that typically, MCI participants show more active effective connectivity or need more effort from stimulus onset to Go indication and then to motion stop, but HC participants show higher maximum and minimum effective connectivity values during the button‐press period, where coherent coordination is required among many region pairs.ConclusionHC and MCI do show significant differences in effective connectivity during each of the three periods of a motion detection task (marked by stimulus onset, Go indication, and button‐press) over many region pairs. Our analysis provides a massive pool of possible features for future biomarker development for effective discrimination of HC and MCI. Funding: NSF‐2032709/Li; NIH‐1R21AG046637‐01A1/Kavcic and NIH‐1R01AG054484‐01A1/Kavcic; NIH‐P30AG053760/Paulson and NIH‐P30AG072931/Paulson.