The language connectome: insights from advanced structural and functional imaging techniques
One of the most intriguing aspects of human kind is our unique capacity for rapidly acquiring speech and language in the early years of life. However, disorders in speech and language are among the most common developmental problems in childhood. They can restrict the child from social participation and academic achievement and may lead to a permanent dysfunction. Equally, children’s intellectual and linguistic or communicative skills play an important role in their evolving sense of competence and self-esteem. Prevention, early diagnosis and improved support of these language disorders is therefore crucial. Language disorders are currently clinically diagnosed using a standardized psychometric test battery measuring different aspects of language processing. However, over the past years, understanding the pathophysiology of these speech and language disorders has gained great interest by neuroscientists. Although a diagnostic imaging marker is at present still lacking, recent functional and structural neuroimaging techniques seem to reveal subtle, but important, functional and/or microstructural changes in the brain of children and adults with language difficulties.
In the last decade, significant progress has been made in more precisely characterising the neurobiology of language in the human brain. The classical language model has been replaced with more comprehensive models influenced by advances in structural and functional neuroimaging methods. These techniques have revealed a far more distributed cortical and subcortical network for the processing and production of language. The emerging framework going forward is one that emphasises processing streams of functional regions anchored by long association fibre pathways and cortico-subcortical projections. However, the current knowledge is far from conclusive and predominantly focused on adulthood. The main goal of this doctoral research is to explore neuroanatomic and neurofunctional substrates underlying language impairment in children as well as in a group of adults after left-hemispheric stroke.
Diffusion Tensor Imaging (DTI) is a non-invasive imaging technique capable of visualising and quantifying the white matter architecture of the brain. DTI provides quantitative metrics of the diffusion process, such as the fractional anisotropy (FA) and apparent diffusion coefficient (ADC), which can be used to evaluate the white matter microstructure. DTI tractography allows to reconstruct long association white matter tracts subserving the functional integration among frontal, parietal and temporal association cortices, needed for the execution of complex language processing tasks. Therefore, this technique qualifies as an adequate imaging technique to investigate a potential neuroanatomic substrate underlying language impairment.
This doctoral research started with a study assessing the structural language connectome in relation to handedness in typically developing school-aged children. The white matter macro- and microstructure was markedly different between left- and right-handed children in both hemispheres. Right-handed children demonstrated a clear left-hemispheric lateralisation of white matter properties for the majority of the investigated tracts, which may reflect the functional dominance of the left hemisphere for language processing. In contrast, a more bilateral pattern of structural lateralisation was revealed for all studied white matter fibre bundles in left-handed children presuming an increased interhemispheric processing of language in these children.
Next, we assessed white matter changes related to language impairment in children with autism, children with developmental dysphasia and children with rolandic epilepsy using DTI. Several structural alterations were found in these children with language impairment. Interestingly, in all these paediatric populations, a structural left-hemisperic dominance of the language network was lacking. In children with autism, children with developmental dysphasia and children with rolandic epilepsy, language lateralisation was more present in bilateral or right-hemispheric language networks. This relative increase of right-hemispheric involvement in language processing might be interpreted as an important compensatory strategy of the central nervous system to stabilise cognitive and in this case specifically language performance.
In a second part of this doctoral research, we aimed to detect why a decreased structural left lateralisation of the language conectome is linked to language problems in children with autism, developmental dysphasia or rolandic epilepsy, and why this is not the case in typically developing left-handed children. As DTI can only provide partial answers on the neurobiological origin of the observed language problems, functional MRI (fMRI) was used to assess the functional connectivity of the language network. fMRI is a technique that allows to indirectly visualise brain activity, both during an active task and in rest, using the vascular response nearby electrically active neurons. First, a active verb generation task was used to localize the cortical regions involved in language function which were then used as seed regions for the study of the resting state connectivity of the brain. Interestingly, the intrinsic functional connectivity did not differ between left- and right-handed children. In contrast, in children with autism, children with developmental dysphasia and children with rolandic epilepsy we found a marked loss of intrinsic functional connectivity. Specifically, their language impairment was linked to a loss of connectivity between the right cerebellum and the supratentorial language areas. These novel findings link the observed language impairment to a relative dissociation of the cortical language system from normal cerebellar control, suggesting that the disturbed development and function of the language system might be due to a loss of normal modulatory control and automation function of the cerebellum.
Finally, we aimed to identify anatomical predictors of language function after left-hemispheric stroke. Our results suggest that DTI based measurements of the dorsal and ventral language stream may serve as clinically useful predictive biomarkers of chronic aphasia after left-hemispheric stroke. Age, gender, lesion size, level of education as well as the DTI derived variables of both the left- and right-hemispheric dorsal and ventral language stream, explained 87% of the variability in language performance of the included patients with chronic aphasia after left-hemispheric stroke.
In conclusion, this doctoral thesis studied the neural correlates of language impairment by applying advanced MR neuroimaging techniques. We have demonstrated several structural and functional changes in the brain of school-aged children and adults with and without language impairment and provided a number of new insights into potential pathways leading to severe language impairment. A better understanding of the neurobiological roots of language difficulties in children, even in utero, may open possibilities for preventive care, early diagnosis and improved support.