SNR and noise measurements for medical imaging. II. Application to fluoroscopic X-ray equipment

Abstract

This p3pr repons on signal-to-noise-rauo (sun) and noise.poruer-specuum (NPS) measurememe on 3 fluoroscopic x-ray unit by the methods of an m l i e r publiccuion I1 is shoen that image nom in fluoroscopy can be constdm,d io bs mulrivxiare n o d and stationary for SmJU m3ge x e s . Therefore, the ided obsener of losi-conmt detvls is the prewhilelung matched filler. This obseser LS sell approximaled by the oc-suppressing obrerter for signals whox spxid frequencies are in a range ahere the .vps i , rewmably constmt In this p3per fluoroscopic imagc noise is chmctenzed by the spaucmmponl >?S. and the noise lag faclor. 3 discnpior of the tempord blnnng of noise. is introduced. In h e fluoroicapic system memmd. I2g is sccn io v q depending on thc lighi inpul to rhe m d x c m c r a md resulrs in "01% I3g factors in the m g e 4.1-9.9 s-!. The dewcubility of an iodine contrast matlend detail in 20 cm m d 30 cm [hick wrylic p h v r o m i is itJdi<d m d SR mrxur"ntc u e ,sed far i ~e i s i n g the dosr-Io-mformmon eomersion cinmncy. S*~R?:D, 01 vlnous fluoroscopic lechniqucs. When the I-n) hcld c o w s the wholc imige inicncificr input held. the use of m mi-scatter gnd resulis to an improvemen! of dose efficiency by n factor of ?-3 as wmpared IO fluoroscopy nithour the gnd. The dose-lo-informauon conversion efficiency tor small x-my fields is seen io be high, and the presence of the mi-sutler grid IS not i m p o m ! an this use. M'hcn aluminum hllmuon IS employcd thc optimum *-ray tube \'olwgc is 5 5 4 0 kV for the 20 cm phantom, and 60-70 kV for the 30 cm phantom P n m w b c a m filtmon by gadolinium (140 m: cm-' Gd202.S) m Xddition Lo d u m n u m (2.S mm AI) R.SUIIP in 3 dorc efficiency impro!ement of up to 40% depending on oiher imaging f3CIOrS The method provides rzsulls of 1m3gc-qualif) mrisureinene on m absoluie scale m d cm dso be used lor performance spenfiutlon or qualiiy conml of fluoroscopic cquipmcnt.

Implementation of a channelized Hotelling observer model to assess image quality of x-ray angiography systems

“…6,10,11 Additionally, application of linear observer models to fluoroscopic and angiographic imaging systems has been investigated. [13][14][15][16][17][18][19][20][21][22][23][24][25][26] However, the vast majority of the reports feature investigations on simulated data. [13][14][15][16][17]19,[21][22][23][24][25] Although simulated data have significant merits for system design, data acquired from real imaging systems are required to assess real-world system performance.…”

“…Only a few reports describe experiments using real images; however, these investigations did not implement channelized Hotelling observer (CHO) models as described below. 18,20,26 For hardware assessment, the Bayesian ideal observer or the ideal linear observer is desirable. 27 When considering object signal-to-noise ratio (SNR) as the model's figure of merit (FOM), the Hotelling model is the optimal linear observer.…”

Efficiency of low‐contrast detail detectability in fluoroscopic imaging

“…These NPS measurements were made from the digitally recorded image sequences described below by the measurement method of Ref. 21.…”

“…21 In the case of temporally white noise, SNR rate 2 would be obtained directly as the product of SNR single frame 2 and the number of frames per unit time: multiplication of SNR rate 2 by the duration of the image sequence would then give the SNR 2 in the sequence of this duration. Here, in the case of temporally colored noise, one cannot proceed that simply; the effective number of statistically independent frames per unit time, F, must be used instead of the actual number of frames.…”

Assessing the Usefulness of the Quasi-Ideal Observer for Quality Control in Fluoroscopy

“…The model observer used in this study has previously been described in detail (8)(9)(10) . In short, The DCsHFs-observer conditional decision variables (eq.1) were needed in (eq.2)…”