Background and purpose Cerebral DSA is currently the gold standard modality for the diagnosis, treatment, and evaluation of many cerebrovascular disorders; however, it remains a considerable source of radiation exposure for patients, a topic of increasing patient and regulatory scrutiny. Modern bi-plane angiography units are developed with the aim of reducing radiation exposure dose while preserving image quality in accordance to the ALARA principle. We present a comparison of radiation dose exposure during diagnostic cerebral DSA performed within one hospital system, utilizing Siemens bi-plane angiography units spanning three different generations.

Materials and methods Diagnostic cerebral DSA studies performed at 4 neuroangiography units at the main hospital and a peripheral hospital from April to November 2014 were retrospectively reviewed. Studies involving imaging of only one vessel or therapeutic interventions were excluded. Four bi-plane rotational fluoroscopy and angiography units (Siemens, Munich, Germany) were involved (from earliest to latest generation): an image intensifier (II)-based Axiom Artis, a digital flat panel (FD)-based Axiom Artis, an FD-based Artis Zee, and an FD-based Artis Q. Up to fifty studies performed on each unit were included. Total radiation dose reported as reference point air kerma (Ka), and total fluoroscopy time were recorded. Total radiation dose was compared across the units using an analysis of covariance (ANCOVA) model, adjusting for body habitus by including patients’ body mass index (BMI) as a covariate. Considering total fluoroscopy time as a surrogate indicator of procedure complexity, average radiation dose over (fluoroscopy) time was also compared in the same manner. Pairwise comparisons were performed with post-test adjustment using the Tukey-Kramer method.

Results 42 to 50 cases were included per bi-plane unit, with a total of 190 diagnostic cerebral DSA studies included in the analyzes. Median patient’s BMI ranged from 26–30 kg/m². Total radiation dose decreased with subsequent newer generations of bi-plane unit: mean Ka of 1525 and 1499 mGy for the earliest II-based and FD-based Axiom Artis units respectively, 1184 mGy for the Artis Zee unit, and 973 mGy for the latest generation Artis Q unit. Radiation dose delivered by the latest generation Artis Q is significantly lower than the earliest generation II-based (p = 0.001) and FD-based (p < 0.001) Axiom Artis units. Average radiation dose over time was highest at 3.0 mGy/sec for the Artis Zee unit, followed by 2.3 and 2.6 mGy/sec for the II-based and FD-based Axiom Artis units respectively. The latest Artis Q unit recorded the lowest average radiation dose, which was significantly lower than the II-based (p = 0.012) and FD-based (p < 0.001) Axiom Artis units, and the Artis Zee unit (p < 0.001). No significant difference in mean radiation dose or average dose over time was observed between the II-based and FD-based Axiom Artis units. Excluding procedures involving 3D angiography did not alter the results.

Conclusion Radiation exposure during diagnostic cerebral DSA studies is reduced significantly with the latest generation of Siemens Artis Q bi-plane angiography unit. No significant difference in radiation exposure is observed between an image intensifier and a flat panel-based system.

Disclosures C. Chung: None. K. Wunderle: None. S. John: None. F. Hui: None.