Abstracts

Rationale: By diffusion tensor imaging (DTI), reduced apparent diffusion coefficient (ADC) and increased fractional anisotropy (FA) have been noticed in epilepsy patients with hippocampal atrophy, cortical dysplasia or other visible brain abnormality. In limited researches from small numbers of patients with MRI-negative epilepsy, similar changes in ADC and FA were also reported. We collected patients with MRI-negative extratemporal lobe epilepsy to evaluate the integrity of white matter. Our aim was to identify the probability density characteristics of ADC and FA in different neural pathways. Methods: 29 healthy volunteers (16 males, mean age = 27±4) and 24 patients with neocortical epilepsy (9 males, mean age = 26±13) were included. DTI data and High-resolution T1-weight fast spoiled gradient recalled echo images (FSPGR) were. Images were first corrected for eddy-current distortion and head motion using FMRIB's diffusion toolbox. Using Analyze 10.0 software, ADC and FA maps were produced. We further extracted ADC and FA and generated nerve fibers for white-matter regions of interest (WM ROIs) separately, each of which was defined in ICBM probabilistic atlas. By DARTEL tool in SPM8, we deformed each ROI from ICBM template to individual with reference to personal SPGR. Then we produced neural fibers by each ROI to calculate lengths of created fibers. Furthermore, we depicted corresponding histograms of FA, ADC and mean fiber length of patient and health groups. Results: Diffusely increased ADC and reduced FA in different pathways were found in the individual analyses and group comparison. All the histograms of ADC, FA and fiber length showed similar shape. The shape of ADC distribution was symmetrical. The ADC frequency peak of patient group was lower and shift to right. The shape of FA distribution was right skewed. The FA frequency peak was higher and shift to left. The shape of fiber length was right skewed. The fiber length frequency peak was higher and shift to left. The differences of distribution shapes were obvious in the posterior thalamic radiation and the sagittal stratum (patient vs. control groups by two-tailed t test, p < 0.001). Conclusions: In this study, we not only confirmed the white matter changes in epilepsy patients without visible brain anomaly but also investigate the probability density of FA, ADC value and mean fiber lengths in detail. Regardless of the epileptogenic zones were determined by EEG in patients with extratemporal lobe epilepsy, common features exist in the major neural pathways. The results indicate that the major neural networks were aberrant and morphological changes in patients with nonlesional epilepsy. The network changes could correspond to specified propagation of epileptic activities and cognitive disorders.