Slide Slam H9
Quantitative MRI reveals differences in subcortical microstructure in children with DLD
Saloni Krishnan1,2, Gabriel J. Cler1, Daniel Papp1, Salomi S. Asaridou1, Kate E. Watkins1; 1University of Oxford, 2Royal Holloway, University of London
Children with developmental language disorder (DLD) struggle to learn their native language for no obvious reason. DLD is a common neurodevelopmental disorder (prevalence ~7%). DLD increases risk for academic underachievement, unemployment, and social and behavioural difficulties. Although we know that DLD does not result from gross neural lesions, we do not have a clear picture of how brain anatomy differs in children with DLD. We previously hypothesised that the dorsal striatum is important for language learning, and may be abnormal in children with DLD (Krishnan, Watkins, & Bishop, 2016). However, empirical evidence supporting this view has been mixed. Factors contributing to the inconsistency of findings across studies include small sample sizes, heterogeneity of groups studied, as well as the scanning methodology used. Here, we report new findings using a robust and cutting-edge quantitative imaging protocol – multiparameter mapping (MPM) – which sheds new light on microstructural neural differences in children with DLD. When using standard structural imaging protocols, e.g., T1-weighted scans, the contrast between grey and white matter reflects a combination of histological properties such as iron content, myelin, cell density and water. Recently, quantitative MRI methods have been used to map specific indices of tissue microstructure, myelination, and macromolecular content, and the resulting maps are highly reproducible across individuals and scanners. For instance, using the MPM semi-quantitative imaging protocol (Weiskopf et al., 2013), multiple maps can be constructed to probe different tissue properties. The generated maps index: 1) the longitudinal relaxation rate R1 (1/T1); 2) the transverse relaxation rate R2* (1/T2*); and 3) Magnetization Transfer Saturation (MTsat). The dominant influence on R1 in cortical tissue is myelin (Lutti et al., 2014), although R1 indexes both myelin and iron in subcortical areas. R2* is sensitive to iron concentration, especially in ferritin-rich regions, such as the basal ganglia, and MTsat indexes myelin through myelin’s interaction with water molecules. The MPM protocol therefore represents an unparalleled means of acquiring time-efficient, multi-modal, whole-brain data with insight into tissue composition. MPM data were collected as part of the Oxford BOLD study (Krishnan et al., 2021). After quality control, we retained data from 56 typically-developing (TD) children and 34 children with DLD. Children with DLD showed lower MTsat values relative to TD children in the left ventral sensorimotor cortex, left Heschl’s gyrus, superior temporal gyrus and caudate nucleus bilaterally. Children with DLD also had lower R1 values across a widespread network of motor, premotor, and temporal cortex, as well as in the caudate nuclei bilaterally. Differences in MTsat and R1 survived whole-brain correction using threshold-free cluster enhancement (P<.05 FWE), and showed overlap in left ventral sensorimotor cortex, superior temporal gyrus and the caudate nucleus bilaterally. No group differences were noted in R2* maps. A voxel-based morphometric analysis of T1w-scans did not reveal any differences, indicating the greater sensitivity to microstructural differences afforded by the use of MPM. These empirical findings strongly support our hypothesis, indicating there are atypical amounts of grey matter myelin in the dorsal striatum and language-relevant cortex in DLD.