From 2013 ASNT Digital Imaging XVI Paper Summaries, July 2013. ISBN: 978-1-57117-242-6.

Copyright © 2013 by The American Society for Nondestructive Testing, Inc.

No part of this document or its contents may be copied, uploaded to the internet, or stored in any shared retrieval system.

Imaging Plate Design and Use for Radiographic Nondestructive Evaluation

Brian S. White
Carestream Health, Inc.
1049 Ridge Road West, Rochester, NY 14615
(585) 627-8017; fax (585) 627-8926; e-mail [email protected]

ABSTRACT
Computed radiography systems require the proper choice of imaging plate for specific applications. Nondestructive
testing requires that imaging plates produce acceptable image quality over a wide range of energy conditions,
material types, material thicknesses, and configurations. The imaging plate type contributes to the overall image
quality and is a key factor in determining productivity. This paper reviews the principles of storage phosphor
imaging plates, and it explores the design contribution towards key image quality parameters such as sharpness,
noise, and brightness. Usage criteria and guidelines for optimum image quality as a function of radiation level and
type will be discussed for the various imaging plate types.

INTRODUCTION
All industrial radiographic systems utilize detectors to absorb radiation to form an image. Computed radiography
phosphor plates capture X-ray and gamma electromagnetic radiation to store an electronic charge. Subsequently
stimulated by a red laser, the plates release blue light proportional to the amount of exposure. This blue light is
collected and converted into a voltage, which is then electronically sampled and processed to form the digital image.

Three types of computed radiography plates currently exist for the industrial market. The plates are segmented by
their resolution capability. However, as discussed in this paper, many attributes in addition to resolution need to be
considered when selecting imaging plate type. Adequate signal, contrast, and low noise are required to detect defects
in an image. Figure 1 presents a graphical representation of image quality for different plate types. The Y axis
represents resolution (sharpness), and the X axis represents the brightness and noise contribution. Sharpness
improves toward the top portion of the graph, while brightness and noise improve toward the right portion of the
graph.

Computed Radiography Plates for Nondestructive Testing

Highest
resolution blue
imaging plates

Relative High resolution

sharpness imaging plates

GeneralFuji ST VI
purpose
imagingCSH
plates
GP

Relative signal to noise

Figure 1: Relative image quality performance for industrial computed radiography plates.

3
General purpose plates have thicker phosphor layers relative to high resolution plates. Blue plates are high resolution
plates that have blue dye added to the phosphor layer to further improve resolution. The image quality
characteristics for each plate type are provided in Table 1. Overall image quality is a function of brightness, noise,
and resolution.

Table 1: Image quality performance for computed radiography plate types.

Classification Brightness Noise Resolution Dose Productivity

general purpose highest lowest lowest 1X highest
high resolution medium low medium 2X medium
high resolution blue lowest highest highest 4X lowest

General purpose imaging plates have the highest brightness, lowest noise, and lowest resolution. Because they have
higher sensitivity, they require less overall dose to achieve an equivalent saturation value. Less time is required to
achieve acceptable images with these plate types. General purpose plates have less resolution and the lowest noise,
relative to the other plate types, with the highest productivity potential.

High resolution imaging plates have medium brightness, low noise, and high resolution. They require twice the
overall dose relative to general purpose plates to achieve an equivalent saturation value, so more exposure time is
required to achieve optimum images with these plates. High resolution plates are significantly sharper relative to
general purpose plate types, with less noise compared to the blue plates.

High resolution blue imaging plates have the lowest brightness, highest noise, and highest resolution. They require
four times the overall dose relative to general purpose plates to achieve an equivalent saturation value. These plates
exhibit the highest noise, and require more exposure time to achieve acceptable images. These tradeoffs are often
considered acceptable to achieve the highest possible resolution for critical applications.

To determine the relative dose requirements for the various plate types, flat-field exposures were made to
approximately 80% of saturation in a 16-bit linear scale. The plates were exposed at 120 kV, 8 mA, at a source-to-
detector distance of 48 inches. The plates were read five minutes after exposure, and their pixel values were
assessed.

It is useful to classify computed radiography systems in terms of image quality metrics and performance levels,
giving the user clearly defined criteria to aid in the selection of imaging plates. Such classifications have been used
for decades with film systems; ASTM E 18151 provides methods for industrial film system classification. Table 2
provides the available industrial film classifications. Film systems have been historically classified by the ratio of
contrast to granularity.

Table 2: Image quality system classification for industrial film systems.

Classification Speed Granularity Resolution

III highest medium lowest
II medium low medium
I low very low high
special very low lowest highest

Imaging plates are classified by signal-to-noise ratio (SNR) (ASTM E 2446).2 The SNR classification is proving to
be inadequate. Table 3 provides the newly proposed classification criteria and limiting values. 3 The SNR
normalized for resolution, the basic spatial resolution of the image, and the equivalent penetrameter sensitivity 4, 5
requirements will become part of the revised system classification standard.

4
 Table 3: Proposed classification standard for industrial computed radiography systems.

Classification SNRn iSRb EPS

CR level III 50 200 1.92
CR level II 70 160 1.66
CR level I 100 100 1.41
CR special 200 50 1.00

Film and imaging plate technology are similar in that both serve as radiation detectors; however, they employ very
different technologies to achieve their final images. For both cases, the exposure levels are proportional to dose.
Digital systems have linear response characteristics; therefore, equal increments of exposure (linear or log scale)
yield a straight-line response of pixel values (Figure 2). The exposure latitude for digital systems is significantly
improved relative to analog film systems. Given the technology differences, film and imaging plates have some
dissimilar design characteristics.

Figure 2: Contrast and exposure latitude relationship between digital and film systems.

THE PRINCIPLES OF IMAGING PLATE DESIGN

The imaging plate type can help determine the achievable image quality. The imaging plate response is influenced
by the phosphor size, phosphor size distribution, and the chemistry of the individual phosphors in the plate. The
imaging plate layer structure, placement, and phosphor thickness also help influence the final image quality
characteristics.

Tradeoffs exist between speed (brightness), granularity (noise), and sharpness. A paradox exists for the plate
designer. With film, as sharpness is improved, the granularity also improves at the expense of speed. With plates, as
the sharpness is improved, the noise gets worse and the brightness is decreased. The overall image quality is
proportional to the speed (brightness), times the sharpness squared, divided by the granularity (noise). The greatest
sensitivity will be achieved with the minimum power to penetrate the specific material thickness and type.

Perceived image quality in computed radiography systems is primarily determined by the contrast-to-noise ratio
(CNR). The ability to perceive wires and penetrameter holes is proportional to the inverse of the square root of the
CNR. Conversely, the SNR is directly correlated to perceptibility. The basic spatial resolution of the image
corresponds to the smallest feature that can be resolved. The ability to perceive objects in a digital image is a
function of dose. Therefore, there is a threshold dose where objects can become acceptably visible in the image.6

5
Table 4 provides an overview of layer thickness for some industrial computed radiography plate types versus
industrial film. Computed radiography phosphor layers can be 10 to 40 times as thick as film emulsion layers. The
additional thickness limits the achievable resolution and noise of the system.

Table 4: Layer thicknesses for some industrial plate types versus industrial film.

Total Overcoat Phosphor Support Backing

Plate (um) (um) (um) (um) (um)
GP 584 11 290 254 29
HR 436 4 160 254 18
XL Blue 436 4 160 254 18

Total Overcoat Emulsion Support Emulsion Overcoat

Film (um) (um) (um) (um) (um) (um)
AA 400 196 1 8 178 8 1

Figure 3 demonstrates these thickness differences with optical cross sections of industrial film versus industrial
computed radiography plates. Because of the additional thickness, plates are more prone to the absorption of
scattered radiation. Scatter occurs when photon energy is absorbed, resulting in lower photon energy (longer
wavelengths) being released. Control and understanding of scatter is very important for computed radiography
imaging plate use, and it is a key difference between plate and film applications. The use of beam collimation and
intensifying screens to absorb low-energy radiation becomes critical for computed radiography applications.

Figure 3: Reflected illumination cross sectional analysis (5000X magnification).

Another key difference between imaging plate phosphors and film emulsions is that phosphor crystals have much
wider particle size distributions. Figure 4 provides an example of one emulsion type versus one phosphor type. The
wider particle size distribution of the phosphors results in additional light scatter. The additional light scatter and
phosphor size further limit the achievable resolution and noise of the overall computed radiography system.

6
 Figure 4: Particle size distribution of silver halide crystals versus phosphor crystals.

Phosphor size, size distribution, and thickness changes are typically employed to influence imaging plate brightness,
noise, and resolution. The composition of the phosphor and plate optics impacts the overall imaging plate brightness.

USAGE CRITERIA AND GUIDELINES

Table 5 provides an overview for imaging plate use as a function of energy level and type. Iridium 192 has its most
dominant output line at 316 keV, with an average energy of 353 keV. Selenium 75 has two dominant output lines at
137 and 265 keV, with an average energy of 217 keV.

Table 5: Imaging plate selection as a function of energy level and energy type.

Radiation Source Energy Type Energy Level Plate

Linear Accelerators X-ray 2–15 MeV GP
Betatron X-ray 2–10 MeV GP
Tubes X-ray > 80 > 220 kVp GP
Tubes X-ray > 80 > 220 kVp HR
Tubes X-ray < 80 kVp HR blue

Cobalt 60 Gamma peaks at 1.17, 1.33 MeV GP

Iridium 192 Gamma seven peaks between 200 - 600 keV HR
Selenium 75 Gamma nine peaks between 66 - 401 keV HR blue

General purpose plates can be used for higher energy applications where the additional imaging plate thickness is
required to absorb the high energy photons. They can also be utilized for applications such as pipe profiling, where
short exposure times are desired. Although general purpose plates have the lowest resolution capability and are more
prone to scattered radiation, they are often favored for gamma radiography, an inherently low-resolution application.
High resolution plates have the most desirable mix of sharpness and noise characteristics. They can be utilized for
typical X-rays or low-energy gamma ray applications such as pipe welds, castings, and aerospace. The very high
resolution blue plates are suitable for applications where the material type has less attenuation. These plates have the
highest resolution and require more dosage to achieve an acceptable image-quality threshold. Typical applications
include electronics, where the features require higher resolution for detection.

7
SUMMARY
The computed radiography plate type that the radiographer selects depends on the application or code sensitivity
requirement. The ability to resolve a particular wire or hole of an image quality indicator depends on many factors.
The final image is part of an overall imaging system that includes exposure technique, imaging plate, hardware,
software, and viewing conditions. The weakest subsystem in the overall imaging system limits the achievable image
quality result. The thoughtful radiographer will select an imaging plate with sharpness and noise characteristics to
provide acceptable image quality; with careful consideration of source type, scatter, and imaging plate absorbance.

ACKNOWLEDGMENTS
The author wishes to acknowledge Steve Mango and Mike Holloway for their review of this paper. Jacques
Roussilhe suggested the topic, and Steve Stello assisted with the particle size analysis. Kathy Delabarto of Eastman
Kodak Company provided the cross sectional analysis.

REFERENCES
1. ASTM E 1815, “Standard Test Method for Classification of Film Systems for Industrial Radiography,” ASTM
International.

2. ASTM E 2446, “Standard Practice for Classification of Computed Radiology Systems,” ASTM International.

3. Ewert, U., ASTM International Meeting, West Conshohocken, Pennsylvania, June 2013.

4. Mango, S., “Qualification of a Computed Radiography System’s Exposure Range for Optimum Image Quality,”
ASNT, Orlando, October 2012.

5. LaCivita, K., and Ewert, U., “Equivalent Penetrameter Sensitivity as Applied to Computed Radiography,”
Materials Evaluation, Volume 70, No. 8, pages 925–930, August 2012.

6. Ewert, U., Heyne, K., Zscherpel, U., Jechow., M, and Bavendiek, K., “Optimum Exposure Conditions for
Computed Radiography Depending On Fixed Pattern Noise and Efficiency of Imaging Plate Scanner Systems,”
QNDE, San Diego, July 2010.