1、Designation: E 2446 05Standard Practice forClassification of Computed Radiology Systems1This standard is issued under the fixed designation E 2446; the number immediately following the designation indicates the year oforiginal adoption or, in the case of revision, the year of last revision. A number
2、 in parentheses indicates the year of last reapproval. Asuperscript epsilon (e) indicates an editorial change since the last revision or reapproval.1. Scope1.1 This practice describes the evaluation and classificationof a computed radiography (CR) system, a particular phosphorimaging plate (IP), sys
3、tem scanner and software, in combina-tion with specified metal screens for industrial radiography. Itis intended to ensure that the evaluation of image quality, as faras this is influenced by the scanner/IP system, meets the needsof users.1.2 The practice defines system tests to be used to classifyt
4、he systems of different suppliers and make them comparablefor users.1.3 The CR system performance is described by signal andnoise parameters. For film systems, the signal is represented bygradient and the noise by granularity. The signal-to-noise ratiois normalized by the basic spatial resolution of
5、 the system andis part of classification. The normalization is given by thescanning aperture of 100 m diameter for the micro-photometer, which is defined in Test Method E 1815 for filmsystem classification. This practice describes how the param-eters shall be measured for CR systems.1.4 The values s
6、tated in SI are to be regarded as thestandard.1.5 This standard does not purport to address all of thesafety concerns, if any, associated with its use. It is theresponsibility of the user of this standard to establish appro-priate safety and health practices and to determine theapplicability of regu
7、latory limitations prior to use.2. Referenced Documents2.1 ASTM Standards:2E 1316 Terminology for Nondestructive TestingE 1815 Test Method for Classification of Film Systems forIndustrial RadiographyE 2002 Practice for Determining Total Image Unsharpnessin RadiologyE 2007 Guide for Computed Radiolog
8、y (PSL Method)E 2033 Practice for Computed Radiology (PSL Method)E 2445 Practice for Qualification of Computed Radiogra-phy Systems3. Terminology3.1 DefinitionsThe definition of terms relating to gamma-and X-radiology, which appear in Terminology E 1316, GuideE 2007, and Practice E 2033, shall apply
9、 to the terms used inthis practice.3.2 Definitions of Terms Specific to This Standard:3.2.1 computed radiology system (CR system)A completesystem of a storage phosphor imaging plate (IP), a correspond-ing read out unit (scanner or reader) and software, whichconverts the information of the IP into a
10、digital image (see alsoGuide E 2007).3.2.2 computed radiology system classA particular groupof storage phosphor imaging plate systems, which is charac-terized by a SNR (signal-to-noise ratio) range shown in Table1 and by a certain unsharpness range in a specified exposurerange.3.2.3 ISO speed SIPxDe
11、fines the speed of a CR system andis calculated from the reciprocal dose value, measured in gray,which is necessary to obtain a specified minimum SNR of a CRsystem.3.2.4 signal-to-noise ratio (SNR)Quotient of mean valueof the linearized signal intensity and standard deviation of thenoise (intensity
12、distribution) at this signal intensity. The SNRdepends on the radiation dose and the CR system properties.3.2.5 modulation transfer function (MTF)The normalizedmagnitude of the Fourier-transform (FT) of the differentiatededge spread function (ESF) of the linearized PSL (photostimulated luminescence)
13、 intensity, measured perpendicular toa sharp edge. MTF describes the contrast transmission as afunction of the object size. In this practice, the MTF charac-terizes the unsharpness of the CR system. This depends on thescanning system itself and IP-type and cassette employed.3.2.6 gain/amplificationO
14、pto-electrical gain setting of thescanning system.3.2.7 linearized signal intensitya numerical signal valueof a picture element (pixel) of the digital image, which isproportional to the radiation dose. The linearized signal inten-sity is zero, if the radiation dose is zero.1This practice is under th
15、e jurisdiction of ASTM Committee E07 on Nonde-structive Testing and is the direct responsibility of Subcommittee E07.01 onRadiology (X and Gamma) Method.Current edition approved June 1, 2005. Published June 2005.2For referenced ASTM standards, visit the ASTM website, www.astm.org, orcontact ASTM Cus
16、tomer Service at serviceastm.org. For Annual Book of ASTMStandards volume information, refer to the standards Document Summary page onthe ASTM website.1Copyright ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959, United States.3.2.8 basic spatial resolutionthe
17、read-out value of un-sharpness measured with duplex wire IQI in accordance withPractice E 2002 divided by 2 as effective pixel size of the CRsystem.4. Significance and Use4.1 There are several factors affecting the quality of a CRimage including the spatial resolution of the IP system,geometrical un
18、sharpness, scatter and contrast sensitivity(signal-to-noise ratio), as well as software. There are severaladditional factors (for example, scanning parameters), whichaffect the accurate reading of images on exposed IPs using anoptical scanner.4.2 This practice is to be used to establish a classifica
19、tion ofCR system classes on the basis of a normalized SNR. Due tothe difference between the methods, it is required to specify theCR system classes with spatial resolution values. The CRsystem classes in this document do not refer to any particularmanufacturers imaging plates. A CR system class resu
20、lts fromthe use of a particular imaging plate together with the exposureconditions, particularly total exposure, the scanner type andsoftware and the scanning parameters. This classificationsystem provides a means to compare differing CR technolo-gies, as is common practice with film systems, which
21、guidesthe user to the appropriate configuration, IP and technique forthe application at hand. The class selected may not match theimaging performance of a corresponding film class due to thedifference in the spatial resolution and scatter sensitivity.Therefore, the practice should always use IQIs fo
22、r proof ofcontrast sensitivity and spatial resolution.4.3 The quality factors can be determined most accuratelyby the tests described in this practice. Some of the system testsrequire special tools, which may not be available in userlaboratories. Simpler tests are described for quality assurancein P
23、ractice E 2445, which are designed for a fast test of thequality of CR systems and long-term stability and are recom-mended as practical user tests, should the user not have thespecial tools available as needed for the tests in this practice.4.4 Manufacturers of industrial CR systems will use thispr
24、actice. Users of industrial CR systems may also perform thetests and measurements outlined in this practice, provided thatthe required test equipment is used and the methodology isstrictly followed. Any alternative methods may be applied ifequivalence to the methods of this practice is proven to the
25、appropriate Cognizant Engineering Organization.4.5 The publication of CR system classes will enablespecifying bodies and contracting parties to agree to particularsystem class, as a first step in arriving at the appropriatesettings of a system, or the selection of a system. Confirmationof necessary
26、image quality shall be achieved by using PracticeE 2033.5. Apparatus5.1 CR system evaluation depends on the combined prop-erties of the phosphor imaging plate (IP) type, the scanner andsoftware used, and the selected scan parameters. Therefore,documentation for each test shall include the IP type, s
27、canner,software and scan parameters, and the results shall be calcu-lated and tabulated prior to arriving at a class assignment. Theapplied test equipment for SNR measurement (Fig. 1) andalgorithm 6.1.1 correspond to Test Method E 1815. The rec-ommended thickness for aperture test object (diaphragm)
28、 is10.2-mm (0.4 in.) of Pb. The SDD shall be at least 1 m (39 in.).TABLE 1 CR System ClassificationCR SystemClassificationMinimumSignal-Noise RatioASTM IP Special/Y 130ASTM IP I/Y 65ASTM IP II/Y 52ASTM IP III/Y 43FIG. 1 Scheme of Experimental Arrangement for the Step Exposure MethodE2446052Do not us
29、e any material (for example, lead) behind the cassetteand leave a free space of at least 1 m (39 in.) behind thecassette.5.2 The step wedge method (Fig. 2) describes a simplerprocedure for SNR measurement than described in TestMethod E 1815, which permits obtaining similar results withless expense,
30、and less accuracy.6. Procedure for quantitative measurement of imagequality parameters6.1 Measurement of the Normalized Signal-to-Noise Ratio(SNR)6.1.1 Step Exposure MethodFor measurement of theSNR, the following steps are taken (see also Test MethodE 1815):6.1.1.1 The IP, with front and back lead s
31、creens of 0.1 mm(0.004 in.) thickness in the typical exposure cassette, shall bepositioned in front of an X-ray tube with tungsten anode. Makethe exposures with an 8 mm (0.32 in.) copper filter at the X-raytube and the kilovoltage set such that the half value layer incopper is 3.5 mm (0.14 in.). The
32、 kilovoltage setting will beapproximately 220 kV.6.1.1.2 Determine the required exact kilovoltage setting bymaking an exposure (or an exposure rate) measurement withthe detector placed at a distance of at least 750 mm (29.5 in.)from the tube target and an 8 mm (0.32 in.) copper filter at thetube. Th
33、en make a second measurement with a total of 11.5mm (0.45 in.) of copper at the tube. These filters should bemade of 99.9 % pure copper.6.1.1.3 Calculate the ratio of the first and second readings. Ifthis ratio is not 2, adjust the kilovoltage up or down and repeatthe measurements until a ratio of 2
34、 (within 5 %) is obtained.Record the setting of kilovoltage for use with the further IPtests.6.1.1.4 The sensitive layer of the IP shall face the X-raysource. For gamma radiography with Ir-192, the measurementsshall be carried out with 0.3 mm lead screens in front andbehind the IP.Also 8 mm Cu shall
35、 be used for pre-filtering (seeFig. 1).6.1.1.5 The scanner shall read with a dynamic range of$12bit and operate at its highest spatial resolution or a spatialresolution for which the classification shall be carried out.Background and anti-shading correction may be used beforethe analysis of data, if
36、 it relates to the standard measurementprocedure for all measurements. The procedure shall be carriedout and documented for all sensitivity and latitude ranges andall read-out pixel sizes if any of these parameters change theSNR-analysis.6.1.1.6 IPs are exposed in a similar way to film radiographyan
37、d under the conditions described: signal and noise (sPSL)orSNR over dose curve shall be measured. It is especiallyimportant that the exposure of the IP for the SNR measure-ments be spatially uniform. Any nonuniformities in X-raytransmission of the cassette front, or defects in the Pb foil or inthe p
38、hosphor itself could influence the SNR measurement. Nomajor scratches or dust shall be visible in the measurementarea. Therefore, exercise considerable care in selection andplacement of the aperture, and selection and maintenance ofthe cassette, the lead screens and the phosphor screen. Toachieve a
39、uniform region of interest on to the IP, the followingstandard protocol is recommended. Other approaches may beused as long as a uniform exposure is created. At least twelveareas (test areas) of $400 mm2(0.62 in.2) are evenly exposedon the same IP over the full working range of dose. Due to thediffe
40、rent construction principles of scanners, the measurementshall be performed for all possible pixel sizes, if the resultschange. The digital read-out intensity values (gray values) shallbe calibrated in such a way, that they are linear in relation to theradiation dose, which corresponds to the photo
41、stimulatedluminescence (PSL) intensity of the exposed IPs. These cali-brated gray values shall be used for the calculation of the SNR.In order to get a reliable result at least six measurements shallbe made on different samples, and the results are to beaveraged for each of the twelve or more dose l
42、evels measured.6.1.1.7 The signal (intensity Imeas) and noise (standarddeviation sPSL) shall be computed from a region withoutshading or artifacts. Sample SNR values shall be taken indifferent regions of the image area under test to ensure thatFIG. 2 Scheme for the Measurement of the SNR by the Step
43、 Wedge MethodE2446053SNR values are within 10 % stable. The size of the ROI usedto measure the mean signal and noise shall be at least 20 by 55pixels and it should be an area ROI. An example technique forassuring reliable signal-to-noise measurements is describedbelow. This can be achieved using a c
44、ommonly available imageprocessing tool. The signal and noise shall be calculated froma data set of 1100 values or more per exposed area. Theunfiltered data set is subdivided into 55 groups or more with 20values per group. For each group with index i, the value Imeas_iis calculated as the mean of the
45、 unfiltered group values and thevalue sPSLiis calculated from the same group values. Anincreased number of groups yields a better (lower) uncertaintyof the result. Due to the filtering effect of this groupingprocedure, the sPSLi-values shall be corrected by the followingequation:sPSLi_corr5 1.0179 s
46、PSLi(1)NOTE 1The values sPSLiare multiplied with 1.0179 to correct for thefollowing median unbiased estimation. Assume k is the number ofconsecutive observations within a group and C is the critical value of thechi-square distribution for a = 0.5 with k-1 degrees of freedom. In case of20 observation
47、s the values sPSLishall be multiplied with 1.0179 forstatistical correction (see also ISO/DIS 10505). The factor 1.0179 corre-sponds to the correction sqrt (k-1)/c) of ISO/WD 10505 for grouping witha group size of 20 elements (k = 20) for application of a median procedure(c = 18.33765).6.1.1.8 The f
48、inal value Imeasis obtained by the median of allImeas_ivalues. The final sPSLvalue is obtained by the medianof all sPSLi_corrvalues. sPSLshall be calculated as referencevalue to a resolution of 100 m, measured with a circularaperture, or 88.6 m measured with a squared aperture. Thefinal value sPSL10
49、0is calculated bysPSL1005sPSL SRmax/88.6! (2)SRmax= Maximum value of basic spatial resolution in mas measured in 6.3.NOTE 2Test Method E 1815 requires the use of a micro-photodensitometer with circular aperture of 100 m* diameter for the measure-ment of granularity sD. Because the pixels in digital images are organizedin squares, the corresponding pixel size is calculated bysqrt (100 m*)2p / 4) = 88.6 mwith: *1m = 3.93701E-05 in.6.1.1.9 The normalized SNR is calculated bySNR 5 Imeas/ sPSL100(3)6.1.2 Step Wedge Method (Manufacturer Test and EnhancedUser Test)T
