ASTM E1935-97(2008)

NOTICE: This standard has either been superseded and replaced by a new version or withdrawn.
Contact ASTM International (www.astm.org) for the latest information
Designation: E1935 − 97(Reapproved 2008)
Standard Test Method for
Calibrating and Measuring CT Density
This standard is issued under the fixed designation E1935; the number immediately following the designation indicates the year of
original adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. A
superscript epsilon (´) indicates an editorial change since the last revision or reapproval.
1. Scope E1441 Guide for Computed Tomography (CT) Imaging
E1570 Practice for Computed Tomographic (CT) Examina-
1.1 This test method covers instruction for determining the
tion
density calibration of X- and γ-ray computed tomography (CT)
systems and for using this information to measure material
3. Terminology
densities from CT images. The calibration is based on an
examination of the CT image of a disk of material with 3.1 Definitions:
3.1.1 The definitions of terms relating to CT, that appear in
embedded specimens of known composition and density. The
measured mean CT values of the known standards are deter- Terminology E1316 and Guide E1441, shall apply to the terms
used in this test method.
mined from an analysis of the image, and their linear attenu-
ationcoefficientsaredeterminedbymultiplyingtheirmeasured
3.2 Definitions of Terms Specific to This Standard:
physical density by their published mass attenuation coeffi-
3.2.1 density calibration—calibration of a CT system for
cient.The density calibration is performed by applying a linear
accurate representation of material densities in examination
regression to the data. Once calibrated, the linear attenuation
objects.
coefficient of an unknown feature in an image can be measured
3.2.2 effective energy—theequivalentmonoenergeticenergy
from a determination of its mean CTvalue. Its density can then
for a polyenergetic CT system. Thus, the actual, polyenergetic
be extracted from a knowledge of its mass attenuation
CTsystem yields the same measured attenuation coefficient for
coefficient, or one representative of the feature.
an examination object as a theoretical, monoenergetic CT
1.2 CT provides an excellent method of nondestructively
system at the effective energy.
measuring density variations, which would be very difficult to
3.2.3 phantom—a part or item being used to calibrate CT
quantify otherwise. Density is inherently a volumetric property
density.
of matter. As the measurement volume shrinks, local material
3.2.4 examination object—a part or specimen being sub-
inhomogeneities become more important; and measured values
jected to CT examination.
will begin to vary about the bulk density value of the material.
1.3 All values are stated in SI units.
4. Basis of Application
1.4 This standard does not purport to address the safety
4.1 The procedure is generic and requires mutual agreement
concerns, if any, associated with its use. It is the responsibility
between purchaser and supplier on many points.
of the user of this standard to establish appropriate safety and
health practices and determine the applicability of regulatory
5. Significance and Use
limitations prior to use.
5.1 This test method allows specification of the density
calibration procedures to be used to calibrate and perform
2. Referenced Documents
material density measurements using CT image data. Such
2.1 ASTM Standards:
measurements can be used to evaluate parts, characterize a
E1316 Terminology for Nondestructive Examinations
particular system, or compare different systems, provided that
observed variations are dominated by true changes in object
density rather than by image artifacts. The specified procedure
This test method is under the jurisdiction of ASTM Committee E07 on
may also be used to determine the effective X-ray energy of a
Nondestructive Testing and is the direct responsibility of Subcommittee E07.01 on
Radiology (X and Gamma) Method.
CT system.
Current edition approved Dec. 15, 2008. Published December 2008. Originally
5.2 The recommended test method is more accurate and less
approved in 1997. Last previous edition approved in 2003 as E1935 - 97(2003).
DOI: 10.1520/E1935-97R08.
susceptible to errors than alternative CT-based approaches,
For referenced ASTM standards, visit the ASTM website, www.astm.org, or
because it takes into account the effective energy of the CT
contact ASTM Customer Service at service@astm.org. For Annual Book of ASTM
system and the energy-dependent effects of the X-ray attenu-
Standards volume information, refer to the standard’s Document Summary page on
the ASTM website. ation process.
Copyright © ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. United States
E1935 − 97 (2008)
FIG. 1 Density Calibration Phantom
5.3 This (or any) test method for measuring density is valid 6.1.3 The physical density of each density standard shall be
only to the extent that observed CT-number variations are determined empirically by weighing and measuring the speci-
reflective of true changes in object density rather than image mens as accurately as possible. It is a good idea to indepen-
artifacts. Artifacts are always present at some level and can dently verify the measured densities using volumetric displace-
masquerade as density variations. Beam hardening artifacts are ment methods.
particularly detrimental. It is the responsibility of the user to
6.1.4 The mass attenuation coefficient, µ/ρ, at the effective
determine or establish, or both, the validity of the density
energy of the system (see 8.3) shall be determined from a
measurements; that is, they are performed in regions of the
reference table. For compounds, µ/ρ can be obtained by taking
image which are not overly influenced by artifacts.
the weighted sum of its constituents, in accordance with the
following equation:
5.4 Linear attenuation and mass attenuation may be mea-
sured in various ways. For a discussion of attenuation and
µ 5 µ/ρ 5 w µ/ρ (1)
~ !
m ( i i
i
attenuation measurement, see Guide E1441 and Practice
E1570.
where:
w = the weight fraction of the ith elemental component.
i
6. Apparatus
6.1.5 For each density standard, the measured density, ρ,
6.1 Unless otherwise agreed upon between the purchaser
shall be multiplied by its corresponding mass attenuation
and supplier, the density calibration phantom shall be con-
coefficient, µ/ρ, as determined in 6.1.4. The linear attenuation
structed as follows (see Fig. 1):
coefficient, µ, thus obtained shall be permanently recorded for
6.1.1 A selection of density standards bracketing the range
each density calibration standard.
of densities of interest shall be chosen. For best results, the
6.1.6 A host disk to hold the density standards shall be
materials should have known composition and should be
fabricated. The opacity of the disk should approximate the
physically homogeneous on a scale comparable to the spatial
attenuation range of the examination objects. If possible, the
resolution of the CT system. It is a good idea to radiographi-
host disk should be of the same material as the examination
cally verify homogeneity and to independently verify chemical
objects, but other requirements take precedence and may
composition. All materials should be manufactured to repro-
dictate the selection of another material.
ducible standards. Solids should be readily machinable and not
susceptible to surface damage. 6.2 In general, it is very difficult to find acceptable materials
6.1.2 One or more cylinders of each density standard shall for density standards. Published density data are generally not
be machined or prepared, or both. Selecting cylinders over reliable enough for calibration purposes. Homogeneity often
rectangles reduces the uncertainties and streaks that sharp varies on a local scale and negatively influences the calibration
corners have on volumetric determination and verification procedure. Machine damage can increase the density at the
methods. The cylinders should be large enough that the mean surface of a sample, making it difficult to determine the density
CT number corresponding to each standard can be computed of the interior material crucial to the calibration process.
overahundredormoreuncorrupted(see8.1.3)pixelsbutsmall Lot-to-lot variations in composition or alloy fraction can make
enough relative to the dimensions of the host disk that radial it difficult to compute mass attenuation coefficients. For these
effects are minimal. and other reasons, development of a good density calibration
E1935 − 97 (2008)
phantom takes effort, resources and a willingness to iterate the 8. Interpretation of Results
selection and production of standards until acceptable results
8.1 Unless otherwise agreed upon between the purchaser
are obtained.
andsupplier,theimageofthedensitycalibrationphantomshall
6.2.1 Liquids make the best standards, because they can be
be analyzed as follows:
precisely controlled and measured. However, liquids require
8.1.1 The phantom scan data shall be reconstructed using
special handling considerations, are sensitive to temperature the same reconstruction parameters and post-processing steps,
variations, and often tend to precipitate, especially high- if any, used for examination object data.
concentration aqueous solutions. It is hard to find organic 8.1.2 The phantom image shall be displayed using the same
display parameters used for viewing examination object im-
liquids with densities above 1.5 g/cm or inorganic liquids
above 4.0 g/cm ; but for many purposes, they offer a suitable ages.
8.1.3 The mean CT numbers of the density standards in the
choice.
CTimage shall be measured. Special attention needs to be paid
6.2.2 Plastics are popular but in general make the worst
to this part of the measurement process.As much of the area of
standards. Most plastics have at best an approximately known
each specimen as practical should be used, but care must be
polymerization and often contain unknown or proprietary
takentoinsurethatonlyvalidpixelsareincluded.Forexample,
additives, making them poor choices for calibration standards.
a square region of interest in a round sample could yield biased
They also tend to vary more than other materials from batch to
results if there are significant radial effects, such as from beam
batch. Notable exceptions to these generalizations are brand-
hardening or a higher density around the perimeter due to
name acrylics and brand-name fluorocarbons.
surface damage caused by machining or compression. Ideally,
6.2.3 Metals are also popular, but they are generally avail-
a circular region of interest should be used that includes a
able only in a limited number of discrete densities. They can
hundred or more pixels but avoids the boundary region around
exhibit important lot-to-lot variations in alloy fractions; but
each density standard, especially if edge effects of any type are
with careful selection or characterization, they can make good
clearly visible.
density calibration standards. Pure elements or very well
8.1.4 A table of linear attenuation coefficients versus mean
known specimens offer an excellent option when they can be
CT numbers shall be prepared.
obtained in the density range of interest.
8.1.5 A least-squares fit to the equation N = a·µ + b shall
CT
6.2.4 Each material must be treated on a case-by-case basis.
be performed on the data stored in the table, where µ is the
Reactor-grade graphite provides a good case study. Reactor-
linear attenuation coefficient and N is the CT number.
CT
grade graphite is available in a variety of shapes, in very pure
8.1.6 The resulting linear curve shall be used as the density
form, and in a number of densities.At first glance, it appears to
calibration. Using the inferred linear relationship between CT
offer an attractive choice in a density range without many
number and linear attenuation coefficient, the measured CT
viable alternatives. However, upon closer examination, the
value, N , of any material can be used to calculate a best
CT
material is found to be susceptible to surface damage during
estimate of its associated linear attenuation coefficient, µ.
machining and to exhibit important inhomogeneities in density
8.2 Unless otherwise agreed upon between the purchaser
on linear scales of about 1 mm. Surface damage makes it
and supplier, the density of a region of interest in an exami-
nearly impossible to determine the core density of the sample
nation object shall be determined as follows:
gravimetrically, because the total weight is biased by a denser
8.2.1 The mean CT number in the region of interest shall be
outer shell. Inhomogeneities make it difficult to extract accu-
measured.
rate mean CT numbers from an image of a sample that is not
8.2.2 From the known calibration parameters, the linear
large in diameter compared to 1 mm.
attenuation coefficient of the region of interest shall be ob-
tained using the equation N = a·µ + b.
CT
7. Procedure
8.2.3 Thedensityoftheregionofinterestshallbecalculated
by dividing the obtained linear attenuation by the appropriate
7.1 Unless otherwise agreed upon between the purchaser
tabulated value of µ/ρ at the effective energy of the system (see
and supplier, the density calibration phantom shall be scanned
8.3). If µ/ρ is not known for the feature of interest, a nominal
as follows:
value for µ/ρ may be used. Variations in µ/ρ are minor, and
7.1.1 The phantom shall be mounted on the CT system with
basically independent of material in the energy range of about
the orientation of its axis of revolution normal to the scan
200 keV to about 2 MeV. Outside this range, the selection of a
plane.
nominal value is more sensitive. Adoption of an appropriate
7.1.2 The phantom shall be placed at the same location used
nominal value is a matter of agreement between purchaser and
for examination object scans.
supplier.
7.1.3 The slice plane shall be adjusted to intercept the
8.3 Unless otherwise agreed upon between the purchaser
phantom approximately midway between the flat faces of the
and supplier, the effective energy of the CT system shall be
disk.
determined as follows:
7.1.4 The phantom shall be scanned using the same data
8.3.1 A table of linear attenuation coefficients versus mean
acquisition parameters, and the data shall be processed using CT numbers shall be prepared for several X-ray energies
the same steps (for example, beam-hardening corrections) bracketing the effective energy of the CT system, as shown in
applied to examination objects. 8.4.1.
E1935 − 97 (2008)
TABLE 2 Density Calibration Data at an Effective Energy of 3800
8.3.2 For each X-ray energy, a least-squares fit to the
keV
equatio
...