Abstract
Dopamine depletion in the axons of Parkinson's disease (PD) patients precedes depletion in cell bodies thus proposing that macroscopic connectivity can be used to understand disease mechanism. A novel multivariate functional connectivity analysis, based on high order coherence among four fMRI BOLD signals was applied on resting-state fMRI data of controls and PD patients (OFF and ON medication states) and unidirectional multiple-region pathways in the sensorimotor system were identified.
Pathways were classified as "preserved" (unaffected by the disease), "damaged" (not observed in patients) and "corrected" (observed in controls and in PD-ON state). The majority of all pathways were feedforward, most of them with the pattern "S1 -> M1 -> SMA." Of these pathways, 67% were "damaged," 28% "preserved," and 5% "corrected." Prefrontal cortex (PFC) afferent and efferent pathways that corresponded to goal directed and habitual activities corresponded to recurrent circuits.
Eighty-one percent of habitual afferent had internal cue (i.e., M1 -> S1 ->), of them 79% were "damaged" and the rest "preserved." All goal-directed afferent had external cue (i.e., S1 -> M1 ->) with third "damaged," third "preserved," and third "corrected." Corrected pathways were initiated in the dorsolateral PFC. Reduced connectivity of the SMA and PFC resulted from reduced sensorimotor afferent to these regions.
Reduced sensorimotor internal cues to the PFC resulted with reduced habitual processes. Levodopa effects were for pathways that started in region reach with dopamine receptors.
This methodology can enrich understudying of PD mechanisms in other (e.g., the default mode network) systems.
