Validation of connectivity-based thalamic segmentation with direct electrophysiologic recordings from human sensory...

Connectivity-based segmentation has been used to identify functional gray matter subregions that are not discernable on conventional magnetic resonance imaging. However, the accuracy and reliability of this technique has only been validated using indirect means. In order to provide direct electrophysiologic validation of connectivity-based thalamic segmentations within human subjects, we assess the correlation of atlas-based thalamic anatomy, connectivity-based thalamic maps, and somatosensory evoked thalamic potentials in two adults with medication-refractory epilepsy who were undergoing intracranial EEG monitoring with intrathalamic depth and subdural cortical strip electrodes. MRI with atlas-derived localization was used to delineate the anatomic boundaries of the ventral posterolateral (VPL) nucleus of the thalamus. Somatosensory evoked potentials with intrathalamic electrodes physiologically identified a discrete region of phase reversal in the ventrolateral thalamus. Finally, DTI was obtained so that probabilistic tractography and connectivity-based segmentation could be performed to correlate the region of thalamus linked to sensory areas of the cortex, namely the postcentral gyrus. We independently utilized these three different methods in a blinded fashion to localize the "sensory" thalamus, demonstrating a high-degree of reproducible correlation between electrophysiologic and connectivity-based maps of the thalamus. This study provides direct electrophysiologic validation of probabilistic tractography-based thalamic segmentation. Importantly, this study provides an electrophysiological basis for using connectivity-based segmentation to further study subcortical anatomy and physiology while also providing the clinical basis for targeting deep brain nuclei with therapeutic stimulation. Finally, these direct recordings from human thalamus confirm early inferences of a sensory thalamic component of the N18 waveform in somatosensory evoked potentials.