The development of epilepsy in the brain involves complex microstructural alterations that are not yet completely understood. Quantitative susceptibility mapping (QSM) based on gradient-echo (GRE) signal phase is uniquely sensitive to local frequency-shifts induced by underlying variations in tissue molecular content that can alter its magnetic micro-environment^1,2. Probing alterations in tissue susceptibility induced by pathology could thus potentially provide more sensitive and specific markers of brain tissue damage than conventional T1/T2/T2* contrasts. Here, we exploit the intrinsic sensitivity of QSM to dia- or paramagnetic alterations in brain tissue to investigate microstructural pathological changes during epileptogenesis in the rat brain. Using the established model of pilocarpine-induced status epilepticus, we show for the first time, detection of localized calcium and iron deposits in specific thalamic nuclei of the epileptogenic brain using QSM.Status epilepticus (SE) was induced in adult Wistar rats by intraperitoneal injection of pilocarpine. At 3-weeks post-injection, SE rats (n=5) and age-matched controls (n=5) were euthanized and brains perfused with 4% PFA. Imaging was performed on an 11.7T MR scanner (Bruker Biospin) using a 20-mm RF transceiver volume coil. GRE data were acquired using a 3D multi-gradient-echo sequence with flip angle 45°, 8 echoes, TE1/ΔTE/TR = 5/5/100 ms, receiver bandwidth = 70 kHz, and isotropic spatial resolution = 120 µm. k-space data were reconstructed in IDL and Fourier-transformed to calculate magnitude and phase images. Parametric R2* maps were generated by exponential least-squares fitting to the multi-echo GRE data. QSM maps were reconstructed from phase data at TE=25 ms. Phase maps were unaliased using 3D Laplacian-based unwrapping³, and spatially-filtered using V-SHARP⁴ (TSVD threshold=0.05, maximum kernel radius=13 voxels) to remove background field contributions. QSM maps were calculated using the TKD method^5,6 (threshold=0.2), and referenced to cortical gray matter for each rat. After MRI, the brains were processed for histology, and serially stained using Alizarin Red (calcium) and Perls’ staining (iron).

Fig. 1 shows comparison of R2* and QSM maps from SE and control rat brains. R2* maps revealed distinct areas in the thalamus which exhibited significantly higher (p<0.001) R2* values in the SE brains (66.5 ± 22.6 s^-1) compared to controls (28.4 ± 4.1 s-1) (Fig. 1A, C). Interestingly, these hyper-intense R2* areas were found to be bilaterally-symmetric (Fig. 1A), and were localized to specific and reproducible regions in the thalamus across the SE cohort examined. In comparison, reconstructed quantitative susceptibility maps at the same level (Fig. 1B, D) revealed significantly heterogeneous contrast within these lesioned areas. Distinct regions within the lesion were found to exhibit significantly (p<0.005) positive (paramagnetic) or significantly (p<0.005) negative (diamagnetic) susceptibility values (as referenced to cortical gray matter) compared to the control brains (Fig. 1B). This heterogeneous contrast can be distinctly seen in the magnified view of QSM maps through the thalamus in Figs. 1B’-D’ (arrows).

Fig. 2 compares R2* and QSM maps from control and SE brains with Perls’- and Alizarin-stained sections from the same rats. Thalamic regions in control brains stained negative for both iron and calcium (Fig. 2C, D), and appeared iso-intense in corresponding R2* and susceptibility maps (Fig. 2A, B). In comparison, we found localized deposits of fine iron- and calcium-positive granules confined to specific dorsal and medial aspects of the thalamus in SE rats (Fig. 2G, H), which corresponded closely to the paramagnetic and diamagnetic regions detected in the QSM maps, respectively (Fig. 2F). High-magnification views of select regions through the SE thalamus (Fig. 2G’-H") depict diffuse microscopic iron and calcium granules deposited in the tissue. Mean susceptibility values measured in lesioned areas corresponding to the panels in Fig. 2G’-H’ and G”-H” were -0.051 ± 0.014 ppm and 0.076 ± 0.027 ppm, respectively, indicating region-specific alterations in tissue susceptibility sensitive to underlying concentrations of calcium and iron observed in these areas. Interestingly, in G”, H” where iron and calcium co-localize, relatively paramagnetic contrast in the QSM map indicated local susceptibility shift dominated by iron concentration (Fig. 2G”).

Calcifications in brain tissue occur as a result of localized increases in calcium concentration in areas of necrotic or apoptotic cell death and inflammation⁷. Our results reveal the potential of QSM to uniquely detect and differentiate between localized deposits of iron and calcium in the epileptogenic rat brain, and further suggest that these depositions occur in specific thalamic nuclei in the SE brains. Histological validation further confirmed that the observed alterations in local susceptibility contrast were dominated by nuclei-specific accumulations of iron and calcium. Previous histological studies have reported calcium granules that grow in size and concentration and crystallize over time after SE^7,8. Investigating the temporal change in QSM contrasts following SE could therefore provide vital insights into the evolving pathology of epilepsy.