Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Patterns of structural variation in human auditory cortex revealed by data-driven curvature-based clustering
Poster Session C, Saturday, September 13, 11:00 am - 12:30 pm, Field House
Yinuo Liu1, Tyler Perrachione1; 1Boston University
The human auditory cortex is distinguished by enormous variation in the patterns of cortical folding seen across individuals. An objective characterization of structural variation in human auditory cortex is important because variation in auditory cortical morphology has been associated with phonological processing (Eckert, 2024), auditory expertise (Golestani et al., 2011), language disorders (Leonard et al., 2001), and language lateralization (Tzourio-Mazoyer et al., 2014). Primary auditory cortex is situated on Heschl’s gyrus (HG), which traditionally has been described as having three distinct gyrification patterns: a single gyrus, a common stem duplication, or a complete posterior duplication (Marie et al., 2015). Situated posterior to HG, gyrification of planum temporale (PT) is also highly variable but less well characterized (Tzourio-Mazoyer and Mazoyer, 2017). Characterization of folding patterns on the superior temporal plane is typically based on manual inspection of brain volumes or surfaces (Morosan et al., 2001; Rademacher et al., 2001), but many patterns of individual morphology are not well described by these three categories. Moreover, prior work has not considered variation in HG and PT folding simultaneously. In this study, we used an unsupervised data-driven clustering approach to characterize folding patterns of the superior temporal plane (including HG and PT) based on a large sample (n=1065) of brain scans from the Human Connectome Project. Using local measures of cortical curvature obtained from FreeSurfer reconstructions, we calculated the Gaussian similarity between all pairs of subjects for gyrification within the superior temporal plane. The resulting similarity matrix was submitted to elbow-optimized k-means clustering to identify common patterns of gyrification across individuals. In contrast to the three classic categories, this data-driven clustering suggested that superior temporal morphology was best characterized by five distinct gyrification patterns. We found two distinct variations of the single HG morphotype: (i) a single HG with evident gyrification of PT (32% of hemispheres) and (ii) a single HG with minimal gyrification of PT (26%). We also found curvature patterns consistent with (iii) classic common stem duplication of HG (16%) and (iv) classic complete posterior duplication of HG (9%). Finally, we observed a previously undescribed curvature pattern intermediate between the classic single HG and common stem duplication, consisting of (v) a wider HG with a notch at the anterolateral end (18%). We also illustrate how these morphological variations are paralleled in variation in the cortical microstructure (intracortical myelination). Overall, these results expand our understanding of the range of individual variation in the morphology of auditory cortex. Future applications of these data-driven techniques may offer new routes to understanding structure-function correspondence in human auditory cortex, including how atypical patterns of gyrification characterize neurodevelopmental or neuropsychiatric variation in language and audition.
Topic Areas: Speech Perception, Methods