Background Three-dimensional digital subtraction angiography (3D-DSA) is a modern technique that allows for better appreciation of complex vascular lesions. This study evaluates the impact of various dose reduction strategies on 3D-DSA image quality.

Methods The standard manufacturer 5 s 0.36 μGy/frame setting was modified to create lower dose 3D-DSA protocols by varying the acquisition time (5 or 3 s) and/or dose per frame (0.36, 0.24, 0.17, and 0.10 μGy/f). All protocols were evaluated in three swine. Four raters measured a segment of the external carotid artery on two-dimensional multiplanar reconstruction images. The raters were also presented with three-dimensional volume rendered images from all protocols in a blinded manner and asked to choose the superior image. A full model analysis of variance with repeated measure factors was performed to compare mean differences in measurements between protocols.

Results Measurement differences between the standard and low dose protocols were not clinically significant (<0.5 mm). All raters demonstrated high inter-rater reliability. The 5 s protocols were considered as qualitatively superior to the 3 s protocols. Delivered system doses ranged from 43.8 to 6.5 mGy. The 5 s 0.10 μGy/frame protocols generated 65–68% less delivered dose compared with the 5 s 0.36 μGy/frame setting.

Conclusions Low dose 3D-DSA protocols with preserved image quality are achievable, and can help reduce unnecessary radiation exposure to both patients and operators. The 5 s low dose protocols generated clinically acceptable and superior images compared with the 3 s protocols, suggesting a more important role for acquisition time than dose per frame to maintain image quality.