WHITE PAPER

Moving from CR to DR
Optimizing Image Quality and Dose
 WHITE PAPER

Table of contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2. CR and DR: similarities and differences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
3. Image capture technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.1 DR panel technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.2 Phosphor technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.3 DR Panel read-out electronics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.4 Pixel fill factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.5 Pixel size (resolution) considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
4. Image processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.1 Image Processing Performance Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.2 What to look for in image processing software . . . . . . . . . . . . . . . . . . . . . . . . 9
5. Anti-scatter grids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
5.1 Grid selection and specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
5.2 Grid line suppression software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
5.3 Grid alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
5.4 Focused grids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
5.5 Grid line orientation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
5.6 Non-grid scatter suppression software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
6. Exposure and dose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
6.1 Dynamic range . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
6.2 The International Exposure Index Standard . . . . . . . . . . . . . . . . . . . . . . . . . . 16
6.3 DICOM mapping of EI, TEI, DI and DAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
6.4 Collimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
6.5 DR repeat rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
7. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
8. Making the move from CR to DR: further reading . . . . . . . . . . . . . . . . . . . . . . . . 19

2 – CR to DR – Optimizing Image Quality and Dose

  WHITE PAPER

1 Introduction
Many imaging departments are making the move from computed radiography (CR) to digital radiography (DR) as
DR is becoming an increasingly attractive choice. While some facilities are in the midst of this conversion, others
are still evaluating their options. In both situations, there are several factors that require attention to ensure a
smooth transition from CR to DR. The technologies do have certain similarities, but often the differences between
them may not be evident until after the CR has been replaced with DR in a facility. The purpose of this paper is to
explain some of these differences and assist imaging providers in managing their migration to DR.

Incremental increases in productivity are critical to help healthcare organizations handle ever-larger
numbers of patients while enhancing patient satisfaction, balancing staffing requirements, and
successfully providing value-based care. The workflow improvements and automation possible with DR
enable hospitals and imaging facilities to care for more patients without increasing staff levels.

Digital radiography also helps facilities move closer to ALARA radiation doses (ALARA = As Low As Reasonably
Achievable) – which is important for all patients, but especially in pediatrics. The dose reduction potential
with DR conversion likewise provides the potential for a reduction in occupational dose for imaging providers.
The improved image quality possible with DR supports greater diagnostic capability and confidence, further
enhancing the role of radiology as the hub of patient care.

From an investment point of view, the cost of DR continues to decrease. As departmental productivity has
increased with DR, the return on investment (ROI) has also improved. Modern DR detectors can be shared
between rooms, devices and operational units, allowing the healthcare enterprise to creatively maximize
investments. However, to fully maximize the DR investment, technologists and facility administrators need
to understand how the technology differs from CR.

This white paper discusses current DR technology and its appropriate utilization to achieve high
image quality at reduced dose while increasing productivity. Topics include DR detector panel
technology, DR image processing, appropriate grid selection, proper panel exposure, exposure
monitoring and X-ray and electronic collimation.

Workflow Improvements: Loma Linda University Medical Center Case Study

DR Retrofit conversion Streamlined

from distributed multi-plate CR workflow
and time savings
Rapid ROI
Loma Linda University
Medical Center (California,
USA) upgraded from CR X-ray
technology to a completely 8.16 minutes 100% FTE gain
less time/exam by reducing
digital system with Agfa
the number of
HealthCare’s Fast Forward DR technologists required
upgrade program. The hospital from two to one on
was able to achieve measurable measured studies
Capacity for 67%
improvements in efficiency and more exams/day;
3,285 more patients/year
patient care as a result of the
upgrade, including:

 CR to DR – Optimizing Image Quality and Dose – 3

  WHITE PAPER

2 CR and DR: similarities and differences 

CR and DR both use a standard X-ray generator. With CR, the X-rays are captured using a photostimulable
phosphor (PSP) plate in a cassette. After the X-ray exposure, the cassette is placed in a special scanner/
reader where the latent image is retrieved pixel by pixel, and digitized. The result is an extremely accurate
image that appears on the CR workstation 30 to 60 seconds after the PSP plate is scanned.

DR captures images directly, using a flat panel detector. DR detectors may be integrated into the
equipment or they may be “cassette sized” (and most often wireless) to work with existing analog X-ray
equipment. DR detectors completely eliminate the step of digitizing the image in a CR reader. Instead, a
scintillator in the DR panel immediately converts the X-rays into visible light that is then converted into a
digital signal (see figure 1 for additional details). The image appears on the DR workstation in just a few
seconds.

Today, there are a broad range of DR units to fit the needs and budget of every imaging facility. Retrofit DR
panels, which come in different sizes and phosphor technologies, enable “Instant DR” by upgrading analog
film or CR-based X-ray systems to DR. DR rooms are available with floor mounted X-ray tubes, ceiling
mounted tubes, and radiography/fluoroscopy (R/F) DR systems, while mobile DR units enable imaging to
be done at the patient bedside in critical care units. Analog mobile units can also be retrofitted with DR to
extend their useful life.

FIGURE 1
DR panel image capture technology

X-ray scintillator screen

Contains phosphors
which convert X-Rays to light Photodiode layer
Collects light and converts
it into electric charges
Electronic control
Triggers the
TFT switches TFT array
Collects charges
from photodiodes

Readout electronics
Capture and amplify
the electronic signals
TFT switching circuit
Analog-to-digital Connects each pixel
conversion to readout device

4 – CR to DR – Optimizing Image Quality and Dose

  WHITE PAPER

3 Image capture technologies

The choice of panel and phosphor technologies can facilitate improved image quality and a dose
reduction of up to 60%, so careful selection is critical.

3.1 DR PANEL TECHNOLOGY

DR uses flat panel detectors to capture images. First, X-rays are absorbed in a phosphor screen layer, inside the
flat panel detector. The X-rays are then converted into visible light. A photodiode converts this light into photo-
charges that are collected via the active matrix TFT sensor of the flat panel detector, creating a signal from each
pixel. These signals are amplified, digitized, processed and sent to the acquisition workstation. From the DR
acquisition station, they are sent to a display, distribution and archival system (commonly referred to as a picture
archive and communication system, or PACS). The digital image can be displayed on a monitor or a hardcopy
image can be printed.

3.2 PHOSPHOR TECHNOLOGY

For both CR and DR, image quality is impacted by the phosphor technology used to convert the X-ray
energy into light. Either powder phosphors (the ‘traditional’ technology) or needle phosphors (the high-
efficiency choice) can be used. CR may use powder phosphors composed of Barium Fluoride Bromide
(BaFBr) or needle phosphors composed of Cesium Bromide (CsBr)1. See figure 2.
For DR, the scintillation layer of the detector can also be made of powder phosphor composed of Gadolinium
Oxysulfide (GOS or Gadox) or needle phosphors composed of Cesium Iodide (CsI). See figure 2.
FIGURE 2
Needle phosphors reduce light scattering and improve image quality
300µ

500µ

CR-BaFBr DR-Csl

To give an example of how the selection of phosphor technology makes a difference: for a high level of X-ray
absorption in the phosphor layer, a thick phosphor layer is needed. When using a powder phosphor, light
scattering in the phosphor layer reduces the sharpness, limiting the image quality. The optimum compromise
of resolution and X-ray absorption is reached for a layer thickness of less than ~300 µm. With a needle
phosphor however, a thicker phosphor layer can be used without jeopardizing the sharpness due to the low
light scattering. With needle phosphor technology, higher X-ray absorption is possible, resulting in lower dose
and better image quality (i.e.) higher detective quantum efficiency or DQE*). A CsI needle phosphor in a DR
system of ~500 µm thickness can have up to 50% more X-ray absorption than a powder phosphor.**

* Detective quantum efficiency (DQE) is a measure of the ability of the imaging system to preserve the signal-to-noise ratio
from the radiation field to the resulting digital image, and thus of the combined effect of sharpness and noise performance.
This value is generally accepted as the best measure of overall performance of imaging detectors in medical radiography.
** CR systems with CsBr needle phosphor plates can achieve similar results, when compared to BaFBr phosphor CR systems.
1
Leblans P.J.R. Vandenbroucke D.A.N., Willems P. “Storage Phosphors for Medical Imaging”, Materials 4(6):1034-
1086 · December 2011. DOI: 10.3390/ma4061034

 CR to DR – Optimizing Image Quality and Dose – 5

 WHITE PAPER

Needle Cesium Iodide (CsI) phosphors technology (figure 2):

   Enable thicker phosphor layers

   Significantly reduce internal light scatter
   Improve image sharpness
   Improve X-ray absorption
   Reduce dose by approximately 50%
   Improve image quality

It is important to note that even though Cesium Iodide detectors normally reduce dose compared to
powder phosphor technology, not all CsI phosphors are the same. Some manufacturers offer less
expensive CsI detectors for the market segments where dose and image quality are secondary to price.
These CsI detectors may use thinner phosphor layers with lower quality phosphors.

3.3 DR PANEL READ-OUT ELECTRONICS

During the read-out process, driver electronics, amplifiers, multiplexers and analog-to-digital convertors
can introduce additional pixel noise. Recent designs of these active components have enhanced noise
management properties, improving image quality at lower doses and allowing for further dose reduction.

3.4 PIXEL FILL FACTOR

Each pixel in a DR detector has an active sensing area (photodiode) and a switching circuit. The “fill factor”
(FF) is the percentage of the pixel that makes up the active sensing area.

Lower quality sensors with larger switching circuits (TFT – thin film transistor) and lines will have lower fill
factors and less active area within each pixel. DR panels with lower fill factors result in lower efficiency and
thus higher image noise and lower overall image quality, compared to DR panels with smaller switching
circuits and better fill factors (see figure 3).

3.5 PIXEL SIZE (RESOLUTION) CONSIDERATIONS

When selecting a DR panel, choosing the appropriate pixel size (sometimes referred to as ‘pixel pitch’) is
important. Factors that should be considered include the type of examination being done (pediatric, adult,
extremities), the dose requirements and the image quality needs.

6 – CR to DR – Optimizing Image Quality and Dose

 WHITE PAPER

While smaller pixel sizes have potential for higher resolution, larger pixel sizes usually have a higher fill factor
and thus a higher collection efficiency. Due to the larger pixel area, more X-ray photons per pixel contribute
to the pixel value, resulting in lower quantum noise. Pixel sizes that are smaller than the resolution
needed for a given examination may actually increase the noise in the image depending on the type of
technology utilized. Using a larger (appropriate) pixel size that still meets the resolution requirements of
the examination can reduce visible noise in the image and offers the potential to lower patient dose.

FIGURE 3
At equal resolution, the smaller switching circuit area (TFT) in pixel B
will result in a larger active area with better DQE and better overall image quality than pixel A

One pixel
One pixel

Photodiode

TFT switch
Row line

Bias line Data line

Scintilator

Pixel A Pixel B
Photodiode
Pixel A = Pixel B in “Resolution”
but not in image quality
Switching circuit
 Switching circuit area of
Pixel A > Pixel B

 Active area of Pixel B > Pixel A

 Thus fill factor and DQE of

Pixel B > Pixel A

 CR to DR – Optimizing Image Quality and Dose – 7

  WHITE PAPER

4 Image processing
4.1 IMAGE PROCESSING PERFORMANCE STUDIES
Studies have demonstrated that image processing can significantly affect perceived image quality at
reduced dose. Multi-scale image processing can improve usable diagnostic information at lower doses2.
(see figure 4). Fractional Multi-scale Processing (FMP) with active noise reduction, provides the potential
for further reductions in dose3.

FIGURE 4
MUSICA multi-scale image processing can improve image quality and reduce dose

Image processing can improve usable diagnostic information at lower dose

Standard 14µGy Multi-scale neonatal

processing processing

Fractional Multiscale Processing (FMP)

FMP is the mathematical substructure of Agfa HealthCare’s latest MUSICA image processing
software, which further decomposes image components into elementary fractions for separate
processing. FMP results in a more accurate multi-scale enhancement model, a balanced
participation of all filter kernel pixels in the enhancement process, and better preservation of
high-resolution, low-contrast details next to high-contrast structures.

2
Sensakovic W.F., O’Dell M.C., Letter H., Kohler N., Rop B., Cook J., Logsdon G., Varich L. Image quality and dose differences
caused by vendor-specific image processing of neonatal radiographs. Pediatr Radiol. 2016 Oct;46(11):1606-13. doi: 10.1007
s00247-016-3663-2. Epub 2016 Aug 3.

3
Vandenbroucke D.A.N., Apgar B.K., Bertens T. Optimizing Patient Dose. Agfa HealthCare White Paper Dec 2014

8 – CR to DR – Optimizing Image Quality and Dose

  WHITE PAPER

4.2 WHAT TO LOOK FOR IN IMAGE PROCESSING SOFTWARE

Image processing software should provide consistent performance, for all body types and patient ages
(neonatal, pediatric, adult, bariatric, etc.), over a wide range of exposure factors. It should not create
artifacts, should be tolerant of over and underexposure, and should especially be low-dose friendly.

The software should increase productivity: reducing, not increasing, work for radiology staff. This means
little to no post-processing, automatic window/level adjustments, automatic electronic masking and
excellent area of interest (AOI) accuracy. Configuration and set-up should be easy: the software should
work well out of the box with little or no ongoing maintenance, include simple and understandable
adjustment settings, and avoid complex parameter adjustments that require set-up and maintenance by
imaging specialists.

5 Anti-scatter grids
5.1 GRID SELECTION AND SPECIFICATIONS
Grid performance factors can also have a significant impact on the digital image. Some of the most
important factors include:

   Grid ratio – impacts image quality

   Bucky factor, or the amount by which the exposure must be increased or decreased when using a grid
– impacts dose

   Reciprocating (moving) versus stationary – impacts visibility of gridlines

(it is important that the gridlines are not visible)

   Line rate (frequency) (in lines/cm or inch) – impacts artifact minimization

   Focal distance – impacts artifact minimization and grid cut off

   Grid positioning, including angle and distance – impacts image quality

   Grid line direction – impacts positioning flexibility

FIGURE 5
Example of anti-scatter grid label displaying specifications

 CR to DR – Optimizing Image Quality and Dose – 9

 WHITE PAPER

5.2 GRID LINE SUPPRESSION SOFTWARE

Most newer DR systems use stationary grids instead of reciprocating grids. To remove grid lines from
the final image, grid line suppression software (GLS) is used. Grid line suppression software automatically
identifies repeating patterns caused by the grid-panel interference, and removes them. The result is
improved viewing conditions and workflow options.

To ensure the grid line suppression software works correctly, it is important to follow the manufacturer’s
recommendations when selecting a grid. The pixel pitch of the panel, the grid type and lines per inch can
influence the appearance of grid lines. Improper grid selection may result in image artifacts caused by
the interference between the DR panel and grid (see images, figure 6). The chart below (figure 7) shows
the results of an image quality evaluation using Agfa HealthCare’s GLS grid line suppression software
in combination with different DR and CR plates and grid specifications. The best results are indicated in
green font.

FIGURE 6

Grid line artifacts Grid line suppression software

(aliasing or moiré patterns) removes artifacts

10 – CR to DR – Optimizing Image Quality and Dose

 WHITE PAPER

FIGURE 7
Agfa HealthCare recommended grid lines/cm for various resolution CR plates and DR Panels

Agfa HealthCare DR Panel Type

Grid Line CR Plate

DR 10S/14S DX-D 30/35C DX-D 40C DX-D 45C
Frequency 100 to 150 micron
32 lines/cm Poor results Poor results Poor results Poor results Poor results
80 lines/inch Not recommended Not recommended Not recommended Not recommended Not recommended

36 lines/cm Poor results Poor results Poor results Poor results Good Results
90 lines/inch Not recommended Not recommended Not recommended Not recommended Acceptable for use

40 lines/cm Good Results Good Results Good Results Good Results Poor results
103 lines/inch Acceptable for use Acceptable for use Acceptable for use Acceptable for use Not recommended

50 lines/cm Best results Best results Best results Best results Better results
132 lines/inch Recommended for use Recommended for use Recommended for use Recommended for use Recommended for use

70 lines/cm Poor results Poor results Poor results Poor results Best results
178 lines/inch Not recommended Not recommended Not recommended Not recommended Recommended for use

80 lines/cm Better results Better results Better results Better results Better results
215 lines/inch Recommended for use Recommended for use Recommended for use Recommended for use Recommended for use

5.3 GRID ALIGNMENT

Incorrect grid alignment can lead to poor image quality. This is a common problem in portable/bedside
imaging and can be the source of many image quality complaints. Correct grid alignment requires
the angle of the tube/collimator to be the same as that of the grid/panel. The tube/collimator should be
parallel to the plane of the anti-scatter grid and the appropriate distance used for the recommended grid
type. Improper alignment will result in increased scatter, grid cut-off and overall poor image quality.

FIGURE 8
Proper tube distance and angle is required relative to the grid used to achieve correct image quality

Distance OK Too high Too low Distance OK Too high

Tube centered Tube centered Tube centered Tube offset Tube offset

 CR to DR – Optimizing Image Quality and Dose – 11

 WHITE PAPER

FIGURE 9 FIGURE 10
Tube collimator and grid are NOT properly aligned (not Tube collimator and grid are properly aligned.
parallel). This results in increased scatter causing poor This results in good scatter removal with excellent
lung and spine detail image detail

5.4 FOCUSED GRIDS

Some grids are labelled as ‘multi-focus’, providing a focal range rather than a single focal distance. This
focal range is always a compromise; the actual focus falls somewhere within the range. At the extremes
of the range, the image will experience grid cut-off and grid lines may be visible at the edges of the image
(see figure 8). Many DR systems offer multiple grid options depending on the distance required for the
examination (Figure 11).

FIGURE 11

For the best results, use the

correct SID and anti-scatter grid
with the right focal distance for 100 cm / 40 inches
the exam.
150 cm / 60 inches
For example: 72 inches for chest
imaging, 40 inches for abdomen,
spine, shoulders etc. 180 cm / 72 inches

12 – CR to DR – Optimizing Image Quality and Dose

 WHITE PAPER

5.5 GRID LINE ORIENTATION

   Most stationary grids used in portable/mobile radiography are the same size as the DR panel being used.

   Grid lines are often oriented along the long dimension, but may be oriented along the short dimension
(aka “decubitus” grids). Decubitus grids may be preferable for chest images done with the detector in
the landscape orientation.

   Grids in tables and upright buckys are usually 17-20” square grids.

FIGURE 12
Grid line orientation

Typical grid line “Decubitus” Square table or upright

orientation grid line orientation bucky grid
“Portrait”

5.6 NON-GRID SCATTER SUPPRESSION SOFTWARE

Although the use of grids is optimal from a physics standpoint, as previously stated, when using a grid
several items must be monitored to ensure proper image quality.

   Grids need to be properly centered and positioned to improve image quality. If they are not,
they can actually reduce image quality and this can easily occur during mobile radiography.

   Grids can become damaged over time.

   Grids may create artifacts in the images (known as aliasing or moiré patterns).

   Grids typically require a higher radiation dose.

   Grids may require longer exposure times.

   Grids add weight to the detector and increase muscle strain for technologists,
especially during bedside/portable radiography.

Because of the time and effort required, in certain cases it may be preferable to avoid using grids,
for example in bedside chest imaging.

 CR to DR – Optimizing Image Quality and Dose – 13

 WHITE PAPER

Recently “non-grid” scatter suppression software has become available which uses advanced image
processing to reduce (not eliminate) the need for a grid. Since scatter radiation is primarily visible in the
low frequency component of an image, scatter suppression software extracts the low frequency scatter
information from the image, while enhancing the medium to high frequencies in the image. This results in
enhancement of the relevant clinical information while reducing the visualization of scatter. For example,
scatter suppression software can improve lung field detail without the use of an anti-scatter grid.

FIGURE 13

Non-grid chest with standard processing Non-grid chest with MUSICA Chest+ software

Scatter suppression software may also be used with a grid in some cases, to provide the best overall
result, for example with bariatric patients. In these cases, high amounts of scatter can be present in the
image even when using a grid, depending on the patient size. Thus, scatter suppression software can
improve image quality.

When using scatter suppression software, each department should establish guidelines for the
appropriate use (or not) of grids, particularly with mobile or portable examinations. The patient type
(pediatric, normal or bariatric) should be considered, as well as the examination criteria (such as ruling out
foreign bodies or providing more prominent visualization of catheters and fine needles) when determining
what type of grid and image processing should be used.

While a grid will increase image quality, grid techniques require proper positioning and a higher overall
dose (up to 50% increase), compared to non-grid exposures.

14 – CR to DR – Optimizing Image Quality and Dose

  WHITE PAPER

In summary, image processing software may be able to eliminate the need for a grid with DR depending on:

   The application: mobile or in-room

   The patient size/type: pediatric, normal, obese, bariatric, etc.

   The workflow requirements: grid + panel weight and correct alignment, SID

   The examination criteria: image quality, dose

Generally, “non-grid” image processing should be seen as an option in the DR “toolbox”. It should be
considered and used when appropriate in order to improve image quality and workflow, and reduce dose.

6 Exposure and dose

6.1 DYNAMIC RANGE
DR images can offer higher contrast and sharpness. DR panels also normally require less exposure (dose)
to achieve equal or better image quality, especially with cesium-based detectors. However, there is
significantly less exposure latitude with DR because of the smaller dynamic range compared to CR.

The dynamic range is the ratio between the highest exposure level without saturation artifacts and the
lowest exposure level detectable above the intrinsic noise level. CR typically has a dynamic range of
approximately 10,000:1. The dynamic range of DR systems is about 200:1 for older systems, and up to
1500:1 for newer, 16-bit systems. Traditional film screen systems have a dynamic range of about 25:1.
Thus, DR systems offer a larger dynamic range than film screen systems, but a much smaller range than
CR systems.

The exposure latitude of a system indicates to what degree the mAs can be changed from the ideal
exposure for a single patient, and still achieve acceptable image quality. The exposure latitude depends
on the type of system used (film, CR or DR), the dynamic range of that system, and the type of body part
being imaged. The acceptable exposure latitude is often expressed as the change in milli-amperage
seconds or mAs, which corresponds to a change in dose.

The exposure latitude of a CR system is much greater than that of a DR or film screen system. For
example, for a lateral skull examination, CR typically has an acceptable exposure range of -4X mAs to
+16X mAs (or more). DR has an acceptable exposure range of about ±4X mAs from the ideal exposure.
A DR exposure variation greater than +4X mAs can result in image saturation, and the data is not usually
recoverable. Therefore, with DR systems, great attention must be paid to exposure accuracy in order to
prevent image saturation and overexposure (see figure 14).

 CR to DR – Optimizing Image Quality and Dose – 15

 WHITE PAPER

FIGURE 14
DR panel saturation at high exposure

CR 70 kVp 160 mAs DR 70 kVp 160 mAs

6.2 THE INTERNATIONAL EXPOSURE INDEX STANDARD

In 2008 the International Electrotechnical Commission (IEC) published the IEC 62494-1 standard,
“Exposure index of digital X-ray imaging systems”. It outlines a methodology for monitoring exposure
consistency within an exam type, and provides a standard way to measure the exposure to a digital
detector.

The index consists of three values: Exposure Index (EI), Target Exposure Index (TEI) and
Deviation Index (DI)4.

   The Exposure Index (EI) is related to the exposure reaching the receptor. If the mAs is doubled, the
exposure index is doubled; if the mAs is halved, the EI is also halved. The relationship is therefore
linear under all conditions. The EI is a relative exposure measurement, within each exam type. It is not
intended (and should not be used) as a calibrated dose meter or a measurement of dose. To confirm the
performance of the exposure index, a flat field exposure can be made with a calibrated (RQA-5) beam. The
IEC standard indicates that the exposure index results should fall within ± 20% under these conditions.

   The Target Exposure Index (TEI) is the reference (e.g. ideal) exposure index for a particular examination
view. Various exams will have different TEI values, depending on the detector type and image quality
needs. For example, the TEI for extremities may be 900; for chest, the TEI may be 250; an abdominal
TEI may be 350; etc. The TEI is used to calculate the Deviation Index (DI).

The goal when selecting the TEI should be to achieve acceptable image quality at the lowest possible
dose (ALARA). Lower TEI values will require less dose, but will reduce image quality. Once a TEI value is
selected, individual EI values may fall outside of these ranges due to normal variations in exposure.

4
Don Steven, Whiting Bruce R., Rutz Lois Jo, Apgar Bruce K. New Exposure Indicators for Digital Radiography Simplified for
Radiologists and Technologists. American Journal of Roentgenology. Dec 2012

16 – CR to DR – Optimizing Image Quality and Dose

 WHITE PAPER

The TEI can be set by an applications specialist with input and recommendations from the hospital, or
can be determined based on the average of a number of exposures. The TEI should never be changed
based on a single image exposure, but rather on statistics from multiple exposures. Check with your
manufacturer for initial TEI recommendations.

FIGURE 15
Agfa HealthCare’s suggested Target Exposure Index (TEI) starting points for various examinations and Agfa
HealthCare DR panel types (CsI or GOS). The final TEI values used should be determined by the imaging
requirements of the radiologists, and normally should be somewhere between the minimum and maximum
values shown.

General work
Examination Shoulders, spines* Extremities
(chest, abdomen, etc)
Detector Starting Starting Starting
Min. Max. Min. Max. Min. Max.
type point point point
DR-Csl
350 150 500 450 200 800 700 400 1000
(needle)
DR-GOS
700 300 1000 900 400 1600 1400 800 2000
(powder)

*If the image quality is acceptable, lower TEI values (100 to 150) may be used on repeated scoliosis studies to minimize dose.

   The Deviation Index (DI) expresses how far a single exposure is from the TEI (the reference). It thus
provides a relative indication of under or overexposure. The DI is equal to 10x the logarithm of the ratio
of EI for a given exam view to the TEI for the exam view. One deviation unit equals ~25% (+1 or -1) over
or underexposure (similar to an automatic exposure control (AEC)). Three deviation units equals 2x or
½x of the target exposure (+3 or -3). While in a perfect world, DI would be 0, this is of course rarely the
case. Studies have shown that when the TEI is properly selected, 90 - 95% of the deviation index values
fall within -3 to +3 deviation units from the target for manual exposures and -2 to +2 deviation units
for AEC exposures. On the Agfa HealthCare DR acquisition station, dose monitoring software must be
active to display DI and TEI values and the associated color-coded graphic on the image.

FIGURE 16
Exposure Index with Deviation Index display

Exposure Monitoring Software

Exposure Monitoring Software includes a color-
coded exposure (dose) bar that indicates how far
the image EI is from the TEI: the dose bar is green
when on/near the target EI, turns yellow when
Red significantly
overexposed over/underexposed by a factor of 2 and red when
over/underexposed by a factor of 4 (see graphic
Overexposed far
display). While the bar gives a relative indication of
outside the range
Greather than
the exposure to the plate, and is a good measure
4X exposure of the variation of the exposure to the plate
Greater than 6.0 DI
within a given exam type, it is not an absolute
measurement of patient dose or absorbed dose.

 CR to DR – Optimizing Image Quality and Dose – 17

  WHITE PAPER

6.3 DICOM MAPPING OF EI, TEI, DI AND DAP

The EI, TEI, DI and DAP values can be displayed on the PACS workstation using the appropriate DICOM
tags. The DICOM committee has defined the following tags for EI, TEI, DI and DAP (Dose Area Product):
   EI: (0018,1411)
   TEI: (0018,1412)
   DI: (0018,1413)
   DAP: (0018,115E)

6.4 COLLIMATION
FIGURE 17
Proper automatic image collimation
X-ray collimation reduces the exposed area, lowering patient dose and
reducing the influence of X-ray scattering. Proper X-ray collimation
significantly impacts dose and image quality with DR. When X-ray
collimation is done correctly, the area of interest should be detected
automatically, and minimal manual cropping should be necessary.

When the X-ray collimated area is larger than the area of interest,
the technologist may wish to electronically crop or mask the image
manually after exposure. However, many facilities have policies in
place that minimize or eliminate this practice, because when cropping
is done the radiologist may be unaware of the actual patient exposure,
both in terms of quantity and anatomy. Therefore, manual cropping or
masking after exposure should be the exception not the rule.

6.5 DR REPEAT RATES

Because of the ease of use, repeat rates with DR are actually increasing in some cases. Repeats are frequently
taken to modify positioning because the image is available quickly and the patient is still positioned on a
detector. While this may offer advantages for the diagnostic and technical quality of images, it should be kept
in mind that this also results in increased radiation dose for the patient. Overall, reduced repeats are a potential
advantage of digital radiography, and care must be taken not to negate these benefits unnecessarily.

 7 Conclusions
To summarize,

1. Moving from CR to DR brings significant improvements in efficiency and patient care

2. Panel technology and phosphor type can reduce dose by 50-60% and influence image quality, so
careful selection is critical.

3. Proper image processing can significantly improve image quality and reduce dose and repeated images
as well. Image processing software should provide consistent performance for all patient profiles, over
a wide range of exposure factors.

18 – CR to DR – Optimizing Image Quality and Dose

  WHITE PAPER

4. Anti-scatter grids and their correct positioning can improve image quality. The required specifications
for grids used for DR may be different than for CR. Therefore, check with your manufacturer.
Depending on the examination, it may be preferable to avoid using anti-scatter grids. “Non-grid” scatter
suppression software can reduce the need for anti-scatter grids in many cases.

5. Proper technique selection is more important than ever. The dynamic range of DR is lower than that of
CR, so DR images can be saturated and unrecoverable in some cases. DR systems should fully conform
to the IEC Exposure Index standard to insure proper monitoring and control of exposure.

6. X-ray collimation influences scatter, image processing and overall image quality. Improper electronic
collimation, cropping or masking to correct for poor X-ray collimation reduces image quality, increases
patient dose, and may be against the healthcare facility policy. Certainly, manual masking/cropping is
not a best practice!

8 Making the move from CR to DR:

further reading
ASRT white paper: Best Practices in Digital Radiography
https://www.asrt.org/docs/default-source/publications/whitepapers/asrt12_bstpracdigradwhp_final.pdf

ASRT article: Radiation Safety Compliance

https://media.asrt.org/pdf/publications/RADT_Vol87_No5_CT.pdf (see page 511 and further)

Agfa HealthCare white paper: Optimizing Patient Dose

http://www.agfahealthcare.com/global/en/he/library/libraryopen?ID=47689266

Agfa HealthCare white paper: Non-grid Bedside Chest Imaging

http://global.agfahealthcare.com/int/non-grid-bedside-chest-imaging-white-paper/

About the authors:

Bruce Apgar is based in Greenville, South Carolina (USA). As Agfa HealthCare’s application lead for imaging
services, he is one of the company’s leading experts on dose reduction issues, especially in neonatal and pediatric
environments. He represents the company, and its views, at several leading technical committees, including
the task groups of the American Association of Physicists in Medicine (AAPM), and at the Medical Imaging and
Technology Alliance (MITA). He has a B.S. in Imaging Science from Rochester Institute of Technology.

George Curley RT(R) has been with Agfa HealthCare for 25 years. He is currently Senior Sales Marketing
Manager of Digital Imaging Products for North America. He is a radiographer and former radiology
manager with extensive experience in digital imaging clinically and commercially.

Dirk Vandenbroucke is an R&D scientist investigating innovative technologies for Agfa HealthCare’s
medical imaging systems. As a senior researcher, he has contributed to the fundamental research
in conventional silver halide film screen systems and in the development of CR and DR systems. He
is an active member of various working groups in international standard committees (ISO, IEC). Dr.
Vandenbroucke has a PhD in physics from the University of Ghent.

 CR to DR – Optimizing Image Quality and Dose – 19

Agfa HealthCare, present in one hospital out of two, is a leading provider of
eHealth & Digital Imaging solutions. Care organizations in over 100 countries
rely on Agfa HealthCare to optimize their efficiency and improve patient care.

Agfa, the Agfa rhombus and MUSICA are trademarks of Agfa-Gevaert N.V., Belgium, or its affiliates. Copyright 2017
All other trademarks are held by their respective owners and are used in an editorial fashion with no intention Agfa HealthCare NV
of infringement. The data in this publication are for illustration purposes only and do not necessarily represent All rights reserved
standards or specifications, which must be met by Agfa HealthCare. All information contained herein is intended
Published by Agfa HealthCare NV
for guidance purposes only, and characteristics of the products and services described in this publication
B-2640 Mortsel – Belgium
can be changed at any time without notice. Products and services may not be available for your local area.
GB 201709
Please contact your local sales representative for availability information. Agfa HealthCare diligently strives to
provide as accurate information as possible, but shall not be responsible for any typographical error.