Mapping Epilepsy-specific Functional MRI Network Properties in Refractory Epilepsy Using Intracranial EEG Locations

Abstract

Epilepsy is a heterogeneous neurological disorder characterized by sudden and unpredictable seizures that disrupt normal brain function. Globally, about 5 million people are diagnosed with epilepsy each year, which makes up a significant portion of the world's disease burden. Epilepsy is increasingly considered a brain network disorder whereby seizure activity in one region of the brain affects the neuronal activity in other brain regions in the network, which means that brain networks are involved in seizure generation and propagation. Brain interactions are known to be dynamic, even in a resting state; therefore, the functional connectivity (FC) of brain regions analysed in fMRI data also varies over time. This thesis aims to study alteration in local dynamic functional connectivity in the regions of interest in the ipsilateral hemisphere (hemisphere where seizures originates), by comparing them to the homologous hemispheric areas. This was done by using the presurgical intracranial electroencephalograpgy (iEEG) electrode implantation as a locator of the presumed epileptogenic zone with rapid functional MRI (fMRI – 608 ms per image) data for dynamic functional connectivity analysis in 17 patients with drug-resistant epilepsy. Dynamic functional connectivity was computed using the dynamic regional phase synchrony (DRePS) method. A comparative analysis using the standard deviation (i.e., variability) of DRePS showed significant temporal variability in ipsilateral regions in 6 out of 17 epilepsy patients. Four showed decreased variability, and 2 showed increased temporal variability in the ipsilateral areas. This indicates an alteration in dynamic functional connectivity in the region of interest in the disease hemisphere. However, non-significant temporal alteration in other subjects might indicate more global network alteration due to disease progression. Large-scale ventricular brain abnormality was also observed in 3 subjects, rendering it challenging to estimate contralateral brain regions in these cases analytically. Understanding the effects of seizure activity on functional connectivity and analysing the properties of network topology can help in further pathological studies and surgical planning in drug-resistant focal epilepsy.