PERCEPTUAL COMPARISON OF PULSED AND CONTINUOUS FLUOROSCOPY

Citation
R. Aufrichtig et al., PERCEPTUAL COMPARISON OF PULSED AND CONTINUOUS FLUOROSCOPY, Medical physics, 21(2), 1994, pp. 245-256
Citations number
36
Categorie Soggetti
Radiology,Nuclear Medicine & Medical Imaging
Journal title
ISSN journal
00942405
Volume
21
Issue
2
Year of publication
1994
Pages
245 - 256
Database
ISI
SICI code
0094-2405(1994)21:2<245:PCOPAC>2.0.ZU;2-5
Abstract
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.