Abstracts

Rationale: Focal cortical dysplasia (FCD) is one of the most common underlying pathologies for medically intractable epilepsies. Although a fair percentage of FCD lesions can be visually appreciated on conventional MRI, it is challenging to distinguish subtle FCD lesions from normal brain tissues, partly due to a lack of sensitive and specific MRI measurements for tissue properties. Recent studies have shown that FCD causes abnormalities in both cyto- and myelo-architectures of the cortex. Here, we investigated quantitative T1 and T2 values as potential biomarkers of tissue properties in epilepsy patients with FCD using a novel high-resolution 3D magnetic resonance fingerprinting (MRF) technique. Prior to the FCD analysis, quantitative T1 and T2 values were investigated in various Brodmann areas to verify the sensitivity of MRF in probing tissue properties of the human cortex. Methods: A 3D whole-brain MRF scan was acquired from 10 healthy subjects and 8 patients with medically intractable epilepsy and histopathologically confirmed FCD (6 FCD type II and 2 MCD, i.e., malformation of cortical development) in a Siemens 3T Prisma scanner (FOV = 300 x 300 x 144 mm3, 1.2 mm3 isotropic voxel size, scan time = 10.4 minutes). The acquired MRF data were interpolated to achieve 0.6 mm3 isotropic voxel size by zero-padding in k-space. T1 and T2 maps were then generated based on the direct matching of the data to a predefined dictionary. T1-weighted (T1w) images were synthesized from the MRF maps and segmented into gray matter (GM), white matter (WM) and cerebral spinal fluid (CSF) after skull extraction using FSL, and then normalized to MNI space using SyN in ANTs. Warping information from the normalization was directly applied to the T1 and T2 maps as they were perfectly registered with the T1w images. The GM masks were applied on the Brodmann atlas (MRIcro) within the MNI space to calculate the average T1 and T2 values for a given Brodmann area. For the FCD cases, a region of interest (ROI) was drawn for each lesion manually. The same ROIs were applied to normal subjects to compare the matrices with FCD patients at the same cortical locations. GM masks were applied on the ROIs to extract the T1 and T2 values of the GM component of the lesions/normal cortices. Results: Figure 1 shows the mean and standard deviation of the T1 and T2 values from 11 Brodmann areas (Brs). The quantitative value changes among different regions are in good agreement with previous literature. Highlighted are the T1 and T2 values in Brs 3 (primary somatosensory cortex) and 17 (primary visual), which are lower than those in Brs 8 and 9 (frontal), corresponding to the different cytoarchitecture of these cortices. Br 29 (retrosplenial) has low T2 value, due to its rich myelin content.MRF results on an example patient with FCD type IIa is shown in Figure 2 upper, with arrows indicating the location of the lesion on the synthesized T1w images. Visual inspection of the T1 and T2 maps shows: (1) on both T1 and T2 maps, the anterior aspect of the right superior temporal gyrus had higher signal intensities compared to the contralateral side (mean GM T1: lesion = 1597 ms and contralateral side = 1567 ms, mean GM T2: lesion = 78 ms and contralateral side = 72 ms); (2) the T1 map in the lesional region had “speckled” appearance, which likely came from higher variations within both the GM and WM components of the lesion (std of GM T1: lesion = 119 ms and contralateral side = 111 ms, std of GM T2: lesion = 8 ms and contralateral side = 7 ms); (3) the synthesized T1w image showed blurring of the GM-WM junction, a typical characteristic for FCD. This patient underwent stereo-EEG which confirmed seizure onset from the lesion location, resection rendered the patient seizure-free (follow up = 2 months). Preliminary analyses from all 8 FCDs showed an increase in T1 of the GM component of the lesions compared to normal controls (Figure 2 lower). When examining FCD subtypes, the T1 increase of the GM component of the lesions was more marked for patients with MCD (T1 14% higher than normal subjects), and less so for patients with type II (T1 3% higher than normal subjects). T2 value of the GM component of the lesions increased slightly on a group level. Further analyses are being performed on the T1 and T2 changes in the WM component of the lesions.  Conclusions: Our preliminary results demonstrate the sensitivity of using multi-parametric MRF results at 3T to differentiate cortical regions with different cyto- or myelo-architecture. We showed the feasibility to use the T1 and T2 values obtained from 3D MRF protocol for characterization of FCD lesions on individual/group levels, although technical and analytical developments are still warranted for this new technique. These findings suggest the potential value of using quantitative MRF protocol as part of epilepsy presurgical evaluation. Funding: This study is supported by NIH R01 NS109439.