Slide Slam R11
Structural connectivity of human inferior colliculus subdivisions using in vivo and post mortem diffusion MRI tractography
Kevin Sitek1, Bharath Chandrasekaran1; 1University of Pittsburgh
Inferior colliculus (IC) is a major computational hub along the ascending auditory pathway that transforms sound to meaning and is purported to play a large role in efferent control of auditory processing. Research in animal models suggests critical roles for the IC in auditory learning, particularly in auditory plasticity, egocentric selection, and noise exclusion. IC contains multiple subdivisions, including central nucleus (receiving ascending inputs) and external and dorsal nuclei (receiving more heterogeneous inputs, including descending and multisensory connections). However, subdivisions of human IC have been challenging to identify using standard brain imaging techniques such as MRI, and the connectivity of each of these subnuclei has not been identified in the human brain. In this study, we estimated the connectivity of human IC subdivisions with diffusion MRI (dMRI) tractography. We used anatomical-based seed analysis as well as data-driven clustering, which provide converging evidence for distinct connectivity profiles for each of the IC subdivisions. Two unique datasets were included in this study. The first is a post mortem human brainstem scanned with dMRI for 208 hours at b=4000 s/mm2 in 120 diffusion directions at 200 µm isotropic resolution. The second is a single in vivo participant scanned over 18 hours at b=1000 and 2500 s/mm2 in 1260 diffusion directions at 760 µm isotropic resolution. Both datasets also included anatomical MRI (post mortem: 50 µm T2-weighted; in vivo: 700 µm T1-weighted and T2-weighted). We conducted two analyses in each dataset. In the data-driven analysis, the entire IC was used as a tractography seed, and the resulting streamlines were partitioned using k-means clustering (k=5). In the seed-based analysis, IC subdivisions were manually labeled on the anatomical images, and tractography was run from each subdivision. In both the post mortem and in vivo datasets, k-means clustering isolated streamlines between IC and cerebral cortex and between IC and brainstem. Each of the clusters reached IC in a unique location: cortex–bound streamlines were centered on dorsal–rostral IC, while brainstem streamlines hit ventrolaterally. Overall, streamline clusters were largely consistent between the post mortem and in vivo datasets. Using anatomically defined IC subdivisions as tractography seeds, streamlines were less differentiated in the in vivo dataset than in the post mortem dataset, where streamlines reaching anatomically defined IC substructures had unique pathways. Compared to central and external nuclei, dorsal nucleus had fewer streamlines connecting caudally with brainstem through lateral lemniscus. However, dorsal nucleus had more streamlines connecting directly with medial geniculate (and beyond towards auditory cortex) than central or external nuclei, agreeing with the data-driven k-means clustering results. In summary, using sub-millimeter diffusion MRI from high quality datasets, we investigated substructure connectivity patterns of human inferior colliculus. With data-driven approaches to cluster white matter connections through IC, we found that streamline clusters reached IC in distinct locations. The results aligned with connectivity patterns based on tractography analysis using anatomically defined IC subdivisions. Taken together, the data-driven and anatomically driven analyses demonstrate that diffusion MRI tractography can reveal fine-grained structural connectivity patterns within the human subcortical auditory system.