Examination of subcortical fMRI causal connectivity patterns in temporal lobe epilepsy
Abstract number :
750
Submission category :
5. Neuro Imaging / 5B. Functional Imaging
Year :
2020
Submission ID :
2423089
Source :
www.aesnet.org
Presentation date :
12/7/2020 9:07:12 AM
Published date :
Nov 21, 2020, 02:24 AM
Authors :
Hernán F.J. González, Vanderbilt University; Saramati Narasimhan - Vanderbilt University Medical Center; Robert Treuting - Vanderbilt University; Graham Johnson - Vanderbilt University; Kristin Wills - Vanderbilt University Medical Center; Catie Chang - V
Rationale:
Temporal lobe epilepsy (TLE) is a focal epilepsy, yet it engenders broad connectivity abnormalities in arousal networks. We previously found that central lateral (CL) thalamic nucleus exhibits abnormal fMRI connectivity with occipital lobe in TLE. Intralaminar thalamic arousal nuclei like centre médian (CM) and CL project to the cortex and basal ganglia, particularly dorsal striatum (caudate and putamen).1 While others have shown disrupted basal ganglia-thalamic connectivity in TLE,2 to our knowledge none have examined connectivity between intralaminar nuclei and basal ganglia. Most resting-state fMRI (rsfMRI) studies utilize connectivity measures that cannot detect directed connections or casual interactions. Here, we applied a novel method of measuring causal connectivity3 of thalamic nuclei (CM and CL) and basal ganglia, and anticipated that causal connectivity of these structures would be decreased in TLE.
Method:
We acquired rsfMRI in 53 patients (29 female, 39±12yrs) with unilateral TLE and 53 controls (29 female, 39±12yrs). FMRI was corrected with RETROICOR and participant-specific brain segmentations were used to generate time series. We calculated causal connectivity of intralaminar nuclei and striatum with the combinedFC metric developed by Sanchez-Romero and Cole.3 CombinedFC between two regions first calculates bivariate correlation between two regions and then partial correlation between these regions using all other regions as confounds. By combining results of bivariate and partial correlation, strength of causal interactions between brain regions is determined.
Results:
Patients with TLE exhibit overall decreased causal connectivity between thalamic nuclei (CM and CL) and dorsal striatum (Fig. 1). In individual nuclei, we observed in patients left CL displayed decreased causal connectivity with right striatum vs controls (p=0.02, t-test). We also found in patients right CM and CL had decreased connectivity with left striatum vs controls (p=0.04 and p=0.01 respectively, t-test). Interestingly, greater causal connectivity between right CM and left striatum (i.e. closer to control values) associated with later age of TLE onset (r=0.3, p=0.029, Pearson correlation). We measured striatal-cortical causal connectivity and found that left striatum in patients had lower causal connectivity with both left parietal and right frontal lobe (p=0.01 and p=0.02 respectively, t-test).
Conclusion:
Patients with TLE exhibit decreased causal interactions among subcortical regions, including between intralaminar thalamic nuclei and dorsal striatum. Disrupted causal interactions among subcortical arousal nuclei may contribute to broad brain network deficits typically seen in patients with TLE. Further work studying causal connectivity in brain networks may improve our understanding of this devastating disorder.
1.Van der Werf et al, Brain Res Rev, 2002;39(2-3):107-40
2.He et al, Brain, 2020;143(1):175-89
3.Romero & Cole, J. of Cog Neur, 2020;0(0):1-15
Funding:
:NIH F31NS106735, R00NS097618, R01NS112252, T32EB021937, T32GM07347, T32EB001628, R01NS110130, R01NS108445.
Neuro Imaging