4.4. Computed Tomography (CT) Interpretation and Cerebral Arterial Calcification Score

The cerebral arterial calcification score was obtained by conducting CT imaging. For analysis of the CT images, all raw Digital Imaging and Communications in Medicine (DICOM) files from the axial reconstruction were downloaded, anonymized, and converted to NifTi format. For each CT volume, a threshold-based technique was used to manually generate regions of interest for calcification within the intracranial vasculature using the open-source 3D Slicer platform version 4.8 [23]. Vessels of interest were the internal carotid, external carotid, and vertebral arteries as well as the basilar artery. A custom Python script was written to determine the total number of voxels containing calcification and converted to physical volume (cm³) through multiplication by voxel size contained in the original DICOM header. All masks were visually inspected for accuracy by a board-certified subspecialist neuroradiologist. Total calcification scores were calculated by summing up the calcification physical volumes in all arterial regions of interest per human brain. The subjects who underwent CT imaging were stratified into tertiles based on their total calcification scores.