抄録
In cerebral angiography for dural arteriovenous fistula （dAVF）, three-dimensional （3D） images obtained from 3D rotational angiography （3DRA） can be especially useful. However, if the contrast injection rates used in 3DRA are inappropriate, poor vessel delineation and reduced visibility of the observed vessels due to contrast reflux are likely to occur. To address this issue, the present study investigated the relationship between the contrast injection conditions for two-dimensional digital subtraction angiography （2D-DSA） and 3DRA, aiming to identify the optimal contrast injection rate for 3DRA based on the contrast injection conditions for 2D-DSA. Cases with dAVF who had undergone both 2D-DSA and 3DRA using subtraction techniques under very similar contrast injection conditions were included. Agreement in the visual evaluation of vessel delineation between angiographies was evaluated using weighted kappa coefficients, while agreement with contrast reflux was assessed using kappa coefficients. Our results showed a good agreement in vessel delineation between 2D-DSA and 3DRA with subtraction, with a weighted kappa coefficient of 0.82. The agreement with or without contrast agent reflux was also good, with a kappa coefficient of 0.85. Our findings indicated that when excellent vessel delineation is achieved in 2D-DSA immediately before 3DRA, clinically useful images can be obtained using the same contrast agent injection rate as in 2D-DSA.
