Abstract
INTRODUCTION: The objective of this article is to evaluate the feasibility of cerebral digital subtraction angiography (DSA) using ultra-low radiation dose settings in a simplified cerebral angiography phantom. MATERIALS AND METHODS: We created a silicone phantom capable of producing a simplified cerebral DSA. A total of 18 DSA sets were obtained with gradual six-step reduction of the detector entrance dose (DED) from 1.82 to 0.08 μGy per frame, while standard, postprocessing algorithm (PPA) and copper filter (0.3 mm) with PPA (CwP) algorithm reconstruction protocols were applied. We quantitatively compared their signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) and qualitatively analyzed the images' qualities in terms of image sharpness, contrast, and noise as investigated by five observers. RESULTS: The SNR and CNR, which decreased with lowering of the DED in the standard protocol group, were significantly compensated by using the PPA. The values were approximately double in the PPA (11.5 ± 2.9) and CwP (11.0 ± 2.5) groups compared with the standard (5.4 ± 1.1) group in the DED of 0.24 μGy per frame as well as in the other values. The total scores of the observers according to the protocols showed a tendency to decrease as the DED lowered. On average, the PPA (96.3 ± 34.6) and CwP (91.3 ± 29.9) groups yielded higher results than the standard protocol (83.7 ± 46.7). CONCLUSION: Given that the current DED ranges from 1.82 to 3.60 μGy per frame for routine cerebral DSA, our results indicate that DED can be decreased to 15%-30% of the current dose level in vessels 2-4 mm in diameter if image-improvement algorithms are applied.