PhD Defence Daniel Uher

Keywords: Focal Epilepsy, Ultra-High Field MRI, MR Image Processing, Functional MRI, PET-MRI 
 

"Advanced image processing in focal epilepsy: Applications of ultra-high field MRI and hybrid PET-MRI"

This thesis investigates the utility of ultra-high field MRI and hybrid PET-MRI along with advanced image processing techniques in patients with 3T-MRI-negative drug-resistant focal epilepsy. The research addresses several key challenges, including a development of a neural network to correct signal inhomogeneities in temporal lobe regions at 7T, particularly for FLAIR sequences, resulting in improved image quality. Furthermore, the concordance between PET-detected glucose hypometabolism and BOLD contrast in resting-state fMRI was examined, revealing reduced BOLD activity in hypometabolic regions in temporal lobe epilepsy patients. Also, the thesis explored thalamic and anterior thalamic nucleus features as potential biomarkers for epileptogenic zone lateralization, finding significant volume differences related to epilepsy characteristics. Finally, applying advanced image processing techniques to a histopathologically-validated focal cortical dysplasia case, demonstrating increased BOLD synchronicity within the epileptogenic zone despite conventional MRI negativity. The research highlights the potential of UHF MRI combined with sophisticated image processing methods to improve detection and delineation of subtle brain abnormalities in drug-resistant focal epilepsy patients. 

Click here for the live stream.