Transition from a detector-dose-based to a CNR-based cerebral angiography system: evaluation of image quality and radiation dose

Abstract To evaluate the image quality and clinical radiation dose of a cerebral angiography system featuring contrast-to-noise ratio (CNR)-based exposure control for neurovascular endovascular treatment (EVT). Image quality was evaluated using a quality control phantom under fluoroscopy and digital subtraction angiography (DSA). Clinical radiation dose was retrospectively compared between a conventional angiography system using detector dose-based exposure control (61 cerebral aneurysm EVT procedures) and a new angiography system using CNR-based exposure control (99 procedures). Patient entrance reference air kerma (K a,r ) and air kerma–area product (P KA ) were analyzed. The new system demonstrated significantly higher observer scores than the conventional system for both fluoroscopy and DSA, including spatial resolution, low-contrast resolution, and dynamic range. The median total K a,r was 1818 [1301–2744] mGy with the new system and 2283 [1580–3184] mGy with the conventional system, representing a 20% reduction (p < 0.05). Conversely, the median total P KA increased by 26% (173 [120–237] vs. 137 [93–207] Gy·cm 2 , p < 0.05), but remained below the current Japanese diagnostic reference level for cerebral aneurysm EVT. Compared with the conventional system, the new angiography platform demonstrated superior image quality under phantom conditions and was associated with a 20% reduction in total K a,r during cerebral aneurysm EVT. Although the total P KA increased by 26%, this increase may have reflected differences in the magnification strategy between the two systems, suggesting a need for further optimization.