The purpose of this study was to investigate the effect on three-dimensiona
l (3D) magnetic resonance digital subtraction angiography (MR DSA) images o
f various Injection protocol parameters (ie, injection order, volume, and r
ate), as well as Image masking. The pelves of 10 normal volunteers were sca
nned using seven different contrast agent volume/injection rate combination
s. Subtraction of a precontrast mask image resulted in vascular image contr
ast improvements of between 4.0 and 7.7 times. Image quality and smaller ve
ssel image contrast In the masked data decreased with increasing injection
number. Data acquired with a high (0.150 mmol kg(-1)) volume yielded the hi
ghest quality images, although only small nonsignificant differences in ima
ge quality and large vessel conspicuity were found between images obtained
using the high and medium (0.075 mmol kg(-1)) volumes. Images acquired with
a low (0.038 mmol kg(-1)) volume, while of lower image contrast, were judg
ed to be of reasonable quality, especially when acquired as the first or se
cond injection. Injection rate (1 ml s(-1), 2 ml s(-1). and 4 ml s(-1)) was
not found to affect the images significantly, although selection of an inj
ection rate that gave an injection duration of similar to 10 seconds tended
to give better vascular Image contrast. Based on these data, a series of e
scalating volumes for multi-injection examination is proposed. (C) 2000 Wil
ey-Liss, Inc.