Identifying the midline thalamus in humans in vivo

Abstract

AbstractThe midline thalamus is essential to flexible cognition, memory, and stress regulation and its dysfunction is associated with several neurological and mental health disorders including Alzheimer’s disease, schizophrenia, and depression. Despite the pervasive role of the midline thalamus in cognition and disease, almost nothing is known about it in humans due to the lack of a rigorous methodology for finding this brain region using noninvasive imaging technologies. Here, we introduce a new method for identifying the midline thalamus in vivo using probabilistic tractography and k-means clustering with diffusion weighted imaging data. This method clusters thalamic voxels based on data-driven cortical and subcortical connectivity profiles, and then segments midline thalamic nuclei based on connectivity profiles established in rodent and macaque tracer studies. Results from two different diffusion weighted imaging sets, including adult data (22-35yrs) from the Human Connectome Project (n=127) and adolescent data (9-14yrs) collected at Florida International University (n=34) showed that this approach reliably classifies midline thalamic clusters. As expected, these clusters were most evident along the dorsal/ventral extent of the third ventricle and primarily connected to the agranular medial prefrontal cortex (e.g., anterior cingulate cortex), nucleus accumbens, and medial temporal lobe regions. The midline thalamic clusters were then bisected based on a human brain atlas (Ding et al., 2016) into a dorsal midline thalamic cluster (paraventricular and paratenial nuclei) and a ventral midline thalamic cluster (rhomboid and reuniens nuclei). This anatomical connectivity-based identification procedure for the midline thalamus offers new opportunities to study this region in vivo in healthy populations and in those with psychiatric and neurological disorders.Significance StatementThe midline thalamus is essential for flexible cognition, decision making, and stress regulation. Abnormal development, dysfunction, and neurodegeneration of the midline thalamus is thought to be a critical pathology in several psychiatric and neurological disorders. Yet, little is known about the role of the human midline thalamus. Here we solve the problem of localization in vivo by using probabilistic tractography and data driven clustering (k-means), a method capable of capitalizing on known connectivity in non-human animal tracer studies. Localizing the midline thalamus based on its connectivity patterns provides a useful tool for future functional and structural human MRI studies to investigate clinical implications of the midline thalamus in diseases such as Alzheimer’s and schizophrenia.