Automatic White Matter Hyperintensity Parcellation: T1w Brain Atlas and Corresponding Software

White matter hyperintensities (WMH) are one of the core pathognomonic correlates of cerebral small vessel disease visible in brain MRI. Here we provide a T1w brain MRI atlas and software for parcellating a given WMH segmentation into 36 anatomical regions of interest (ROI) following the previously established bullseye parcellation of WMH [1] (which effectively captures all visual rating scales for WMH). This technique first divides white matter into 4 equidistant concentric layers extending from the ventricles to the cortex. Secondly, frontal, parietal, temporal and occipital lobes are delineated for both hemispheres of the brain, while the basal ganglia, thalami and inter-lobe regions are combined (known as the BGIT region). We provide a brain atlas template for obtaining the lobar segmentation via registration, along with software to construct the concertric layers using output from the SynthSeg[2] tool. This tool allows for a rapid and robust registration based parcellation. We do not provide a WMH segmentation tool here, instead our parcellation tool may be used in conjunction with any WMH segmentation algorithm, or for parcellating any other ROI into the same anatomical regions of interest. For the latest version of our pipeline, see https://github.com/BenjaminPhi5/ auto-wmh-bullseye-parc. Our software pipeline has been designed and tested for Linux OS command line use only, however it should work on any platform where FreeSurfer can be successfully installed (basic requirements for FreeSurfer can be found here: https://surfer.nmr.mgh.harvard.edu/fswiki/SystemRequirements).