Presentation

Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Poster Slams

Hemispheric asymmetries in the cortical myeloarchitecture parallel the functional lateralization of language

Poster E50 in Poster Session E, Saturday, October 8, 3:15 - 5:00 pm EDT, Millennium Hall
Also presenting in Poster Slam E, Saturday, October 8, 3:00 - 3:15 pm EDT, Regency Ballroom

Tyler Perrachione¹, Ja Young Choi²; ¹Boston University, ²Harvard University

Left-hemisphere dominance for language is the oldest and best-established example of functional specificity in the human brain. Even 150 years after the foundational discoveries of Broca and Wernicke, exactly why language occupies primarily the left frontal and temporal lobes remains an enduring question. Asymmetries in gross morphological features like gray matter volume or cortical surface area remain distantly removed from the neuronal-level tissue properties that could explain hemispheric asymmetries in neural computations. Advances in structural MRI now allow better analysis of signals related to local tissue cytoarchitecture (Glasser & Van Essen, 2011), providing more proximal resolution on the microstructural differences that could inform this question. Here, we used structural and functional MRI scans from the Human Connectome Project (HCP) Young Adult 1200 Subjects dataset to measure (1) intracortical myelination, (2) cortical thickness, and (3) functional selectivity for language within each anatomical region of the Desikan-Killiany Atlas from FreeSurfer. Using T1- and T2-weighted anatomical volumes, we obtained the T1w/T2w ratio—a measure of tissue microstructure sensitive to local differences in intracortical myelination (Glasser & Van Essen, 2011). We measured cortical thickness from the T1-weighted volume. We used the HCP Language Processing fMRI task (Binder et al., 2011), which contrasts neural responses to auditory stories (Story condition) vs. auditory arithmetic problems (Math condition) to measure language selectivity, which we operationalized as the proportion of voxels within a region having greater Story vs. Math response among those with positive response to the Story condition. For each dependent measure, we computed its regional lateralization index (λ) as the difference between left and right hemisphere values divided by their sum: λ=(L-R)/(L+R). Cortical myeloarchitecture asymmetries: We observed a striking anterior-posterior gradient in intracortical myelination. Frontal and temporal regions exhibited significant leftward lateralization, while parietal and occipital regions exhibited significant rightward lateralization. The most left-lateralized regions were IFG pars triangularis (λ=0.041), rostral MFG (λ=0.039), and IFG pars opercularis (λ=0.038) (all p<<0.0001). Temporal areas were also significantly left-lateralized: STG (λ=0.022); MTG (λ=0.016). The most right-lateralized regions included lingual (λ=−0.033), pericalcarine (λ=−0.030), and inferior parietal cortices (λ=-0.030). Language selectivity asymmetries: We found strong leftward asymmetries in language-selective responses, with the strongest in IFG pars opercularis (λ=0.433), caudal MFG (λ=0.419), and IFG pars triangularis (λ=0.315). Lateralization was significant but weaker in MTG (λ=0.173) and STG (λ=0.042), reflecting more bilateral organization of speech processing. Relationships between cortical myeloarchitecture and language selectivity: Five core language regions exhibited significant correspondences between individual subjects’ language and myeloarchitectural asymmetries (controlling for age and sex; Holm-Bonferroni correction for multiple comparisons): STG (β=0.173, p<0.0001); IFG pars orbitalis (β=0.121, p<0.0025); IFG pars triangularis (β=0.098, p<0.05); IFG pars opercularis (β=0.092, p<0.05) and MTG (β=0.088, p<0.05). No such relationships for cortical thickness: Cortical thickness exhibited a subtle but significant pattern of predominantly rightward lateralization across the brain. However, there were no relationships between cortical thickness and language lateralization. Together, these results suggest that left-hemisphere dominance for language may be related to hemispheric asymmetries in the cytoarchitecture of core language regions—especially IFG and STG—favoring faster, intracortical myelin-dependent computational circuitry.

Topic Areas: History of the Neurobiology of Language, Methods