Pulsed fluoroscopy (hereafter called pulsed) at reduced acquisition ra
tes, typically 15 acq/s (pulsed-15), is proposed to reduce x-ray dose
in interventional procedures. However, since the human visual system (
HVS) acts as a temporal low-pass filter that interacts with such acqui
sitions, the proper dose for pulsed must be obtained in perception exp
eriments. We determine the dose for low-frame-rate pulsed that gives v
isualization equivalent to that of conventional 30 acq/s fluoroscopy,
hereafter called continuous. Computer-generated phantoms are used. The
y consist of stationary, low-contrast disks on a Aat background contai
ning Poisson noise that mimics quantum noise in fluoroscopy. Image seq
uences are displayed on the video tachistoscope, a device with conside
rable display flexibility. Three experimental paradigms are used. (1)
In a paired-comparison study, pulsed and continuous are displayed side
-by-side on the same monitor, and the visibility of a contrast detail
phantom is compared. (2) Using this same display, subjects record the
minimally detectable disk contrast (the min-contrast measurement). (3)
In a four-alternative forced-choice experiment, a disk is placed in o
ne of four positions, and the subject determines the position of the d
isk. The methods are complementary-the forced-choice experiment proper
ly eliminates the subjectivity of the observer threshold while the pai
red-comparison study is much more time efficient. With regard to pulse
d and continuous comparisons, remarkable similarity is found between t
he supra-threshold experiments (1 and 2) and the detectability experim
ent (3); i.e., the average absolute differences in the equivalent-perc
eption dose as determined by the three measures is approximately 3%. N
o difference is found between interlaced and noninterlaced display. A.
relatively small dependence of dose savings on disk size is found wit
h larger disks giving increased dose savings. Average dose savings of
22%, 38%, and 49% are found for pulsed-15, pulsed-10, and pulsed-7.5,
respectively.