Abstracts

DECREASED REGIONAL HOMOGENEITY, A MEASURE OF LOCAL FUNCTIONAL CONNECTIVITY, IN INTRACTABLE FOCAL EPILEPSY

Abstract number : 2.166
Submission category : 5. Neuro Imaging
Year : 2012
Submission ID : 16361
Source : www.aesnet.org
Presentation date : 11/30/2012 12:00:00 AM
Published date : Sep 6, 2012, 12:16 PM

Authors :
K. Weaver, J. G. Ojemann, A. Poliakov, N. Kleinhans, G. Pauley, T. Grabowski, E. Novotny,

Rationale: Functional connectivity (fc) imaging identifies patterns of brain connectivity by correlating intrinsic, low-frequency fluctuations of BOLD signal during task-free, resting state fMRI. Regional homogeneity (ReHo) is a new fc approach that was developed to estimate the magnitude of connectivity within a local environment (i.e. the degree to which the time-course within a given voxel is associated with nearest neighbors). In this pilot study, we employed ReHo to examine the degree to which fc is altered within the local cortical environment immediately adjacent to and surrounding the seizure focus Methods: Four pre-surgical patients with intractable, focal epilepsy (EP; age range 18-37, 2 females) and 4 matched controls (MC) underwent resting state fMRI scans (3T, TR/TE: 2s/21ms, volumes: 160). Standard pre-processing pipelines were employed to remove non-neuronal sources of variance including 1) motion realignment, 2) spatial smoothing 3) grand-mean intensity normalization, 4) linear drift and CSF signal removal and low-pass temporal filtering of frequencies below 0.01 Hz (cf Power et al., NeuroImage 59:2142,2011). Data were then passed through a Kendal's coefficient of concordance ReHo (KCC-ReHo) analysis, computing an average correlation coefficient across the time domain for each (center) voxel with its surrounding 27 neighboring voxels. Each 3D ReHo data set was then normalized by the global mean KCC-ReHo value within the brain mask. Finally, images were spatially registered and sub-divided into regions-of-interest (ROIs) as outlined by the MNI atlas (fig 1a). A mean KCC-ReHo value was calculated for each ROI and compared across groups using a paired-sample student's T test for three regions: 1) seizure focus, 2) an uninvolved region located in the occipital pole and 3) whole brain value. Seizure focus was confirmed from the post-operative resection following invasive monitoring. Results: The Figure (a) shows 5 ROIs surrounding the seizure focus for one patient. The difference between KCC ReHo values between the patient and control are greatest in the ROI that houses the focus (basal temporal, *-p<.05). The KCC ReHo differed significantly between patients and controls for both the whole brain and the focus (Figure 1b; *-p<.05). For the seizure focus ROI , mean KCC-ReHo was lower for EP than MC individuals (EP: 2.47+/-0.29; MC: 3.54+/-0.27, p<0004), and whole-brain normalized KCC-ReHo was also lower (EP: 3.83+/-0.29; MC: 4.07+/-.25, p=0.02) but with no difference in the uninvolved (occipital lobe) region (EP: 3.73+/-.27, MC: 4.07+/-.14, p=0.24). Conclusions: These preliminary results suggest a disruption in local functional connectivity and potentially a desychronization of the regional low frequency, resting state fMRI signal at the seizure focus in focal epilepsy. Future work will explore the degree to which ReHo is associated with post-operative outcomes and its capacity to facilitate identification of the seizure foci in focal epilepsy.
Neuroimaging