This head phantom dataset comprises 6.978 axial images of head / neck CTA phantom AVM 51-04 acquired with a Canon Aquilion ONE scanner.
The data is organized in 16 series with a dose of 10.5 mGy and 8 reconstruction methods: Filtered back projection (FBP), standard hybrid iterative reconstruction (AIDR 3D), and deep-learning reconstruction (AiCE). FBP and AIDR 3D images come with three kernels (FC08, FC13, FC35) and AiCE comes with the "standard" and "bone" kernel. Each acquisition is available with a slice thickenss of 1mm and 0.5mm.
The data provides comprehensive coverage of anatomy and pathology (arteriovenous malformation of the right hemispher) across various reconstruction methods in state-of-the-art CT imaging.