Neurosurgery Resident Schulich School of Medicine London, Ontario, Canada
Rationale: The thalamus is intricately associated with the hippocampus (Jones, 2007). It consistently demonstrates atrophy and connectivity-related changes in patients with temporal lobe epilepsy (TLE) (Li et al., 2012; Santyr et al., 2018). Thalamic nuclei-specific study can improve our understanding of hippocampal-thalamic connections in TLE patients and potentially improve outcomes in neuromodulation therapies. Despite frequently targeting thalamic nuclei in stereotactic neurosurgery, individual thalamic nuclei cannot be directly visualized on routine clinical magnetic resonance imaging (MRI) sequences. Therefore, targeting techniques are indirect, relying on stereotactic coordinate systems, histological atlases, and electrophysiological recording. These systems do not account for inter-subject variability. Higher field MRI affords the opportunity to directly visualize the thalamic nuclei in vivo, allowing for individualized targeting and analysis of disease-related changes.
Methods: In a series of healthy individuals and patients with TLE, T1-weighted, T2-weighted, and MP2RAGE images were acquired using 7 Tesla (7T) MRI at a resolution of 0.7mm3. To obtain a high-resolution composite image, these were registered across subjects, averaged together, and transformed to the common AC-PC space. Representative slices of the thalamus were chosen for segmentation and comparison with the Schaltenbrand-Wahren and Morel atlases. Three thalamic nuclei closely integrated in seizure propagation, the anterior nucleus of the thalamus (ATN), mammillothalamic tract (MTT), and centromedian nucleus (CM) were manually segmented in a subset of five healthy controls and five TLE patients for comparison of volume and mean T1map values. Results: Imaging at 7T in healthy individuals provides sufficient resolution and contrast within the thalamus for the visual identification of the anterior, centromedian, ventralis oralis, ventralis intermedius, ventralis caudalis, medial dorsal, and pulvinar nuclei and mammillothalamic tract (Figure 1). In the small subset of subjects examined, there was no significant difference (p >0.05) in volume or mean T1map between TLE patients and healthy controls for the three thalamic nuclei of interest: ATN, MTT, and CM (Figure 2).
Conclusions: Ultra-high field MRI at 7T provides a method of direct visualization of thalamic nuclei, uncovering substructures not previously identifiable in vivo. These advances will enable quantitative analysis of disease-related changes to these structures and improved clinical targeting. Initial ‘proof-of-concept’ subset analysis of thalamic nuclei implicated in TLE yielded no significant differences in volume or T1map between TLE patients and healthy controls. The consistent tissue characteristics and visible appearance demonstrated here reinforces the reliability of these targets for further study. As such, quantitative diseases-related metrics are currently being explored further with a larger number of subjects and automated segmentation of these structures of interest.