The use of proton MR spectroscopic imaging and MRI volumetric measurements in the clinical evaluation of partial epilepsies

The goals of the research work presented in this thesis were (i) to determine the accuracy of lateralization of temporal lobe epilepsy (TLE) by proton magnetic resonance spectroscopic imaging (MRSI) and magnetic resonance imaging (MRI) based volumetric measurements (MRIVol) in comparison with the current "gold standard" of clinical-EEG lateralization, (ii) to optimize the sensitivity of the MRSI and MRI methods (iii) to examine the surgical outcome in groups of patients undergoing surgical treatment, and (iv) to provide insights into the pathogenesis of partial epilepsies.;Early on we confirmed the usefulness of measuring hippocampal formation (HF) volumes in the lateralization of TLE, and showed, for the first time, that the additional determination of amygdala (AM) volumes improves lateralization (paper 1). These findings were specific for temporal lobe epilepsy (TLE), as patients with extra-temporal partial seizures did not have atrophy of the AM or HF (paper 1). In several subsequent original studies we focused on the pathophysiology and clinical manifestations of TLE, including the relationship of mesial temporal atrophy (MTA) to seizure frequency and duration (paper 6); the association of MTA with early childhood prolonged febrile convulsions (paper 7); the relation of MTA to scalp (papers 4 and 5) and intracranial EEG findings (paper 3); the relationship between atrophy of the amygdala and ictal fear in TLE (paper 9); the frequency and characteristics of dual pathology (paper 8); the correlation of MTA with surgical outcome (paper 10) and the question of qualitative vs. quantitative assessment of MTA (paper 2).;The work on MRSI included the development of a new protocol for assessment of neuronal metabolic abnormality in vivo in TLE based on the N-acetylaspartate (NAA) resonance measured in anatomically defined regions of the temporal lobes (paper 11). Subsequently, we were the first to show the advantages of combining proton MRSI and MRIVol for lateralization of TLE (paper 12). We also demonstrated that MRSI could be used to discriminate between different types of nonconvulsive seizures (paper 15) and to monitor the progression of chronic encephalitis and focal seizures (Rasmussen's syndrome) (paper 14). More recently we began developing new protocols for the use of MRSI in the lateralization and localization of non-lesional extra-temporal epilepsies (paper 13). (Abstract shortened by UMI.).