Objective:

To study the effect of unilateral cortical spreading depolarizations (CSD) on resting state functional connectivity (RSFC).

Background:

CSDs are implicated in the pathophysiology of several neurological conditions including traumatic brain injury (TBI). Previous work has suggested that CSDs induced in one hemisphere increase intrahemispheric RSFC and decrease interhemispheric RSFC using hemodynamic measures of RSFC. We sought to determine the effect of CSDs on neuronally-derived RSFC in a mouse model and apply our analysis approach to a patient with lateralizing TBI.

Design/Methods:

We utilized Thy1-ChR2 mice to non-invasively induce unilateral CSDs with optogenetics in mice with chronic, intact skull imaging windows. We then determined hemodynamically-based RSFC at baseline, immediately following CSD, and at 1, 12, and 24 hours post-CSD using optical intrinsic signal imaging under tribromoethanol anesthesia. We performed neuronal calcium imaging in a separate cohort of Thy1-GCaMP6f animals to determine neuronally-based RSFC after a CSD. We used global regression and a novel local-regression approach where the CSD hemisphere was regressed separately from the contralateral hemisphere. Finally, we analyzed fMRIs from a patient with lateralizing TBI and a healthy control patient using analogous regression approaches.

Results:

CSD disrupted hemodynamic and neuronal metrics of RSFC. Furthermore, we found that this effect was exaggerated by global signal regression. We developed a novel local regression processing technique (hemispheric regression) which mitigated these differences. Our results showed differences between these two signal processing techniques. Finally, we applied the hemispheric regression approach to a patient with lateralizing TBI and found analogous differences.

Conclusions:

We found considerable dynamic changes after CSD both in neuronal and hemodynamic-derived RSFC. Furthermore, a local hemispheric regression approach led to differences in connectivity patterns compared to the commonly used global signal regression.