アイテム詳細

要旨

OBJECTIVE: The superior colliculus (SC) plays an important role in representing the external environment across multiple sensory modalities. Ultra-high field MRI enables imaging and characterization of SC with a greater signal-to-noise ratio in smaller voxels. We investigate the role of deeper layers of SC in somatovisual integration (SVI) that uses tactile cues to guide a visual task. These responses are compared to a visual-only (VO) version of the same task. METHODS: MRI data were collected at 9.4T using a 16 (31)-channel transmit (receive) array. T1-weighted anatomical images were acquired with an MP2RAGE sequence (0.6 mm isotropic voxels). fMRI data were collected with 1-mm cubic voxels and 1.5-s TR (8 runs). A level-set scheme (Truong et al., Hum Brain Mapp 41, 2020) was used to quantify depth-dependence of the fMRI signal. Surfaces of each SC were divided into a 3x3 grid to allow across-subject averaging. Data were bootstrapped across subjects to quantify its reliability. Eight individuals participated in SVI and VO sessions, at least one each. Air puffs were delivered to their fingers during the SVI task to cue them to attend to a quadrant of the visual field. Subjects reported the total number of ‘+’ patterns that appeared in this quadrant after each 6-min run, while ignoring other patterns. Single puffs were alternately presented to the index and ring fingers; a random double puff cued visual attention to the upper (index finger stimulation) or lower (ring finger stimulation) visual fields. Stimulation alternated between the left and right hands (cuing left and right visual fields) every 15 seconds, enabling coherence analysis. During the VO task, the target visual quadrant was cued by a color change without tactile stimulation.
RESULTS: We found that somatosensory stimulation drives lateralized responses in SC fairly globally. In rostral SC, responses were largely superficial. Deeper lateralized responses were found in caudal SC, as expected for forelimb stimulation. Compared to the VO responses, deep caudal SC responses were stronger in the SVI task, reaching significance at some depths. However, the SVI responses were noisier than those evoked by VO, and additional evaluation of noise sources is underway. CONCLUSION: We probed multisensory responses and studied the depth dependence of sensory processing within human SC using 9.4T fMRI, demonstrating its utility for investigating sensory responses within human midbrain structures. Consistent with prior animal studies, results suggest that the deeper layers of the SC represent multisensory information in a retinotopic fashion.