Altered Hemodynamic Oscillations Of Resting-State Networks In Mesial Temporal Lobe Epilepsy

Xiaopeng Song,Yi Zhang,Hang Zhang,Qijun Xie,Shuyan Song,Haibo Xu,Yijun Liu

DOI: https://doi.org/10.1117/12.2006958

2013-01-01

Abstract:Mesial-temporal lobe epilepsy (mTLE), a neurological disorder characterized by abnormal synchronous discharges in a large cell population, affects the hemodynamic activities of functional networks remote from the epileptogenic zone and causes widespread deficits in brain functions. Although a number of resting-state fMRI studies have found altered spatial patterns in the canonical resting-state networks (RSNs) in patients with mTLE, including the default mode network (DMN), dorsal lateral attention network (DAN), auditory network (AUN), somatosensory network (SMN) and visual network (VIN), none of these studies has addressed the question whether the frequencies of hemodynamic oscillations in these RSNs were altered. In the present study, we have proposed a network-based temporal clustering analysis (TCA) method to characterize the resting hemodynamic activity of a large-scale functional network. First, the RSNs were identified in healthy controls as well in the left mTLE patients using independent component analysis (ICA). Then, a time course representing the hemodynamic activity of each RSN was extracted by counting the number of the voxels that were activated simultaneously at each time point within the network. Finally, the power spectral density (PSD) of the time course was estimated. Our results have demonstrated significant differences in the frequency profiles of the SMN, VIN and left DAN between the patients and controls: the peaks of these spectra shifted toward a lower frequency in the patients, while more power was distributed over higher frequency bands in the healthy controls. However, no significant difference has been found in the AUN, DMN and right DAN. These features might serve as biomarkers to differentiate the patients from controls.