The vascular organization of the human brain can determine neurological and neurophysiological functions, yet thus far it has not been comprehensively mapped. Aging and diseases such as dementia are known to be associated with changes to the vasculature and normative data could help detect these vascular changes in neuroimaging studies. Furthermore, given the well-known impact of venous vessels on the blood oxygen level dependent (BOLD) signal, information about the common location of veins could help detect biases in existing datasets. In this work, a quantitative atlas of the venous vasculature using quantitative susceptibility maps (QSM) acquired with a 0.6 mm isotropic resolution is presented. The Venous Neuroanatomy (VENAT) atlas was created from 5 repeated 7 Tesla MRI measurements in young and healthy volunteers (n = 20, 10 females, mean age = 25.1 ± 2.5 years) using a two-step registration method on 3D segmentations of the venous vasculature. This cerebral vein atlas includes the average vessel location, diameter (mean: 0.84 ± 0.33 mm) and curvature (0.11 ± 0.05 mm -1 ) from all participants and provides an in vivo measure of the angio-architectonic organization of the human brain and its variability. This atlas can be used as a basis to understand changes in the vasculature during aging and neurodegeneration, as well as vascular and physiological effects in neuroimaging.

The vascular organization of the human brain can determine neurological and neurophysiological functions, yet thus far it has not been comprehensively mapped. Aging and diseases such as dementia are known to be associated with changes to the vasculature and normative data could help detect these vascular changes in neuroimaging studies. Furthermore, given the well-known impact of venous vessels on the blood oxygen level dependent (BOLD) signal, information about the common location of veins could help detect biases in existing datasets. In this work, a quantitative atlas of the venous vasculature using quantitative susceptibility maps (QSM) acquired with a 0.6 mm isotropic resolution is presented. The Venous Neuroanatomy (VENAT) atlas was created from 5 repeated 7 Tesla MRI measurements in young and healthy volunteers (n = 20, 10 females, mean age = 25.1 ± 2.5 years) using a two-step registration method on 3D segmentations of the venous vasculature. This cerebral vein atlas includes the average vessel location, diameter (mean:0.84 ± 0.33 mm) and curvature (0.11 ± 0.05 mm -1 ) from all participants and provides an in vivo measure of the angio-architectonic organization of the human brain and its variability.This atlas can be used as a basis to understand changes in the vasculature during aging and neurodegeneration, as well as vascular and physiological effects in neuroimaging. Keywords: venous vasculature; QSM, UHF-MRI; vein atlas; vein segmentation; cerebral vasculature information about venous properties, which we seek to leverage here. In this paper we introduce the Venous Neuroanatomy (VENAT) atlas. This atlas provides several advantages over existing venous atlases, since it was acquired with a higher field strength (higher SNR), has a higher image resolution (0.6 mm isotropic) and uses multiple repetitions to increase SNR. This image resolution allows detection of vessels up to 0.3 mm in diameter. TheVENAT atlas is open source and available on figshare (https://figshare.com/articles/VENAT_Probability_map_nii_gz/7205960). This atlas was built using an optimized approach through registration of vascular segmentations and using five repeated QSM images per participant. The VENAT atlas includes a partial volume (PV) map, which shows vessel location, maps of the diameter, curvature of vessels, and venous density maps.