Samuel Lapere1, Evelien Carrette2,
Paul Boon3, Kristl Vonck3, Xavier De Tige4,
Els Fieremans5, Ali Tabesh5, Eric Achten1,
Karel Deblaere1
1Department of Radiology,
Ghent University Hospital, Ghent, Belgium; 2Reference Centre for
Refractory Epilepsy , Department of Neurology, Ghent University Hospital,
Ghent, Belgium; 3Reference Centre for Refractory Epilepsy,
Department of Neurology, Ghent University Hospital, Ghent, Belgium; 4Laboratoire
de Cartographie Fonctionnelle du Cerveau, ULB-Hpital Erasme, Brussels,
Belgium; 5Center for Biomedical Imaging, Department of Radiology,
New York University School of Medicine, New York, NY 10016, United States
In patients with refractory epilepsy, diffusion tensor imaging (DTI) can non-invasively provide valuable information on the microstructure and architecture of brain tissue in vivo and can be used to detect and evaluate microstructural alterations of white matter, even beyond the visually abnormal area. However, DTI has an important limitation because it incorrectly assumes that water diffusion in biological tissues occurs in an unrestricted manner and follows a Gaussian distribution, in which the diffusion weighted signal attenuatestion occurs monoexponentially with the strength and duration of the diffusion gradient (i.e. b-value). Diffusion kurtosis imaging (DKI) is an extension of the DTI technique which takes the deviation of restricted water diffusion in biological tissues from the Gaussian distribution (i.e. excess kurtosis characterizing the non-monoexponential signal decay) into account, thereby offering a more sensitive method of detecting subtle microstructural changes in neural tissue, both in the predominantly anisotropic white matter and the more isotropic grey matter . In this study, DKI was used in patients with refractory epilepsy with unknown seizure focus to examine both the region indicated by a preceding magnetoencephalography (MEG) exam (i.e. magnetic dipole cluster) and, if present, the grey matter abnormality visible on anatomical MRI scans.
Keywords