1996, The British Journal of Radiology, 69, 563-566

Contrast reduction using energy dependent

intensifying screens
G A SHAH, G HASSAM and D L NEWMAN

Department of Radiologic Sciences, Faculty of Allied Health Sciences and Nursing, Kuwait University, Kuwait

Abstract
The speed of intensifying screens, in particular rare earth screens, increases with beam energy. Using such a
screen, the measured contrast of an aluminium disc, with optical densities in the linear region of the
characteristic curve, was found to be significantly and consistently less than that calculated from the gamma
of the film-screen combination for all tube potentials from 40 to 140 kVp. This contrast reduction effect can
be explained in terms of the filtering action of the disc which increases the effective energy of the beam.
Hence, the film-screen combination is exposed to X-rays which have different effective energies. This results
in two characteristic curves of different speed. The contrast of the disc image is not dependent on the gamma
of either curve, but an effective gamma which links the two curves. This effective gamma is less than the
gamma of either curve. Contrast calculated using the effective gamma gave good agreement with the measured
contrast of the disc image over the tube potential range examined.

The energy dependent attenuation of a heterogenous
X-ray beam leads to beam hardening as the beam passes
through an attenuating medium. Therefore, in conven-
tional radiography the beam transmitted through ana-
tomical structures and onto the screen-film system has
a higher effective energy than the incident beam. Rare
earth as well as calcium tungstate screens exhibit signi-
ficant energy dependency; speed increases with beam
energy [ 1 - 5 ] . However, this effect is more marked for
rare earth screens [5].
A recent observation has highlighted the fact that these
energy dependent processes might interact in such a way
that would lead to a reduction in the expected contrast.
the exposures Ei and E 2 are of different effective energies
since the X-ray beam passing through the anatomical
structure has been hardened, i.e. effective energy of Ex is
greater than the effective energy of E 2 . As the effective
energy of the beam increases, the shape of the character-
istic curve remains essentially unchanged but it shifts to
the left [ 7 ] . This leads to the situation illustrated in
Figure 2. Exposure E 2 operates on curve A and exposure
Ei on curve B. Therefore, instead of a contrast C given
by OD 2 —OD 1; the resultant contrast C is given by
OD 2 — ODi'. As a consequence, the contrast achieved is
less than would be expected from the value of y. In effect

Theoretical considerations
If we consider the film-screen response in terms of the
characteristic curve shown in Figure 1, then gamma of
the system (slope of the characteristic curve in the linear
region) is given by the following equation [6]:
£ OD2
Gamma = (OD 2 -OD 1 )/(log 1 0 E 2 - l o g 1 0 E t ) (1)
In the absence of scatter, the radiographic contrast, C, Q
of a structure of thickness t and attenuation coefficient O
pi is given by [6]: O

C = Gamma x subject contrast C = 0.4343yfit (2)

"5.
OD,
where y is the gamma of the system.
However, when imaging an anatomical structure
which has a higher attenuation than its surroundings,

Received 4 December 1995 and in revised form 13 February

1996, accepted 26 February 1996. LogE, LogE2
Address correspondence to Dr G A Shah, Medical Log Exposure
Radiation Technology, Faculty of Medicine and Health
Sciences, University of Newcastle, NSW 2308, Australia. Figure 1. Typical characteristic curve.

Vol. 69, No. 822 563

 G A Shah, G Hassam and D L Newman

Attenuation coefficient
At each tube potential the half value layer (HVL) was
determined using type 1100 aluminium sheets and
narrow beam geometry, as shown in Figure 3. The linear
attenuation coefficient was calculated for each tube
potential using the following equation [6]:
C = 0.693 (3)
d)
a
"o
o Results
Characteristic curves of the Lanex/T Mat G system
O with no additional aluminium in the beam for different
OD, tube potentials are shown in Figure 4. As may be seen,
the range of exposures chosen was sufficiently large to
cover the gamma region of the film. The energy depen-
dency of the screen may be clearly seen, the speed increas-
ing by a factor of 3.4 (at an optical density of 1.0 above
base plus fog) between 40 kVp and 140 kVp. The effect
log E, log E2 on the film response of the additional 2 mm of aluminium
Log Exposure is illustrated in Figure 5, which shows the characteristic
curves obtained at 140 kVp with and without the alu-
Figure 2. Schematic shift of the characteristic curve for the minium. It can be seen that a substantial shift in response
attenuated beam to the left, y and y' defined. occurs as a result of beam hardening. A similar shift was

the gamma of the system is reduced to a value y' as

illustrated in Figure 2.

Methods
A gadolinium oxysulphide rare earth screen (Lanex
Detail) was used in combination with T-Mat G Film
(Kodak, Rochester, USA). All exposures were made with
FILTERS
a high frequency X-ray unit (Picker Clinix E, Cleveland,
USA). The inherent filtration of the X-ray tube was
0.5 mm Al equivalent and the added filtration was
2.5 mm Al. The film was processed using a Kodak RP
X-Omat, model M6B processor operating at 35 °C and
a cycle time of 90 s.
Disc contrast
A 2 mm aluminium disc was placed on the film-screen 200cm
cassette and imaged at each tube potential. Optical den-
sity values OD 2 and ODj were measured with a densi-
tometer, X-Rite model no. 301X (X-Rite, Grandville,
USA).
Gamma of the screen-film
The characteristic curves of the screen-film combi-
nation were determined from 40 to 140 kVp in 20 kVp DETECTOR
steps. Tube current was kept constant; only exposure
time was varied to obtain the required optical density.
Exposures were measured with a Victoreen 660-1 elec- LEAD SHIELD
trometer and a Victoreen model 660-3 ionization BACKINGv
chamber (Victoreen, Cleveland, USA). The detector res-
ponse is linear over the range of exposures measured
and its sensitivity is essentially unchanged for photon
energies from 30 to 140 keV. Two characteristic curves FLOOR
were obtained at each tube potential with and without Figure 3. Schematic diagram of the experimental arrangement
an additional 2 mm aluminium sheet placed in the beam. used to determine the half value layer.

564 The British Journal of Radiology, June 1996

Contrast reduction using intensifying screens

Table I also shows C calculated using y' (vide supra),

which gives reasonable agreement with the measured
contrast. The last column, C'/Cm, gives values which
vary around 1.00, the value for a perfect fit; the average
being 1.01. Thus, it can be seen that contrast is reduced
because the effective gamma is lower than that derived
from the characteristic curves.

Discussion and conclusions

The speed of a given film-screen combination increases
with the effective energy of the beam [ 1 - 5 ] . As a conse-
Log Exposure quence of this, the contrast of an anatomical structure
which attenuates the beam so as to increase its effective
Figure 4. Influence of tube potential on the characteristic curve energy is less than would be calculated using the gamma
of film-screen combination investigated. The curves shift to the of the film-screen system. However, if one incorporates
left as tube potential increases. the shift of the characteristic curve to determine a new
gamma, called y', then there is reasonable agreement
o - between the calculated and measured contrast.
—•— 140 (2mm Al)
—•— 140 (no Al)
It may be seen (Figure 4) that there is only a small
P
increase in screen speed with tube potentials for tube
Opt ica 1 Density

potentials greater than 100; however, a noticeable shift

in speed still occurs when the beam is hardened
/
(Figure 5). This apparent inconsistency may be due to
Y differences in the beam hardening mechanism resulting
from either a tube potential increase or filtration. In the
case of beam hardening by an increase in tube potential
1_ 1
higher energy X-rays are produced whilst in beam fil-
tration, beam hardening results from a selective removal
Log Exposure of the lower energy components of the beam.
The reduction in contrast was less at 40kVp (15%)
Figure 5. Characteristic curves at 140 kVp with and without a than for higher tube potentials (30%). At 40kVp the
2 mm thick aluminium filter. contrast is greatest and there is less latitude. As a result
the density values ODi and OD 2 may lie where curves
observed at the other tube potentials employed. From A and B begin to converge, i.e. close to the inflection
the measured values O D t ' at each tube potential the points in the curves (Figure 2). Then y' will become
corresponding value of ODj was obtained from each set similar to y and the contrast reduction effect would be
of response curves by measuring the optical density cor- reduced. It follows, therefore, that this effect will only be
responding to log10 Ei on curve A. The gamma values significant for low or moderate contrast structures in
y and y' were then derived in the manner indicated which the densities lie within the gamma region of the
in Figure 2. characteristic curve. While these investigations have been
Table I shows the values of HVL, \i, y, C (contrast conducted without a scattering phantom in the beam,
calculated using Equation (2)) and Cm (measured con- the relevance of this effect will also be modulated in
trast) over the range 40-140 kVp. It can be seen that Cm clinical practice by the presence of scatter which will
is consistently less than the theoretical contrast. The lead to further contrast reduction.
reduction is approximately 30% for tube potentials The effect of beam quality on speed is apparently less
greater than 40 kVp while at 40 kVp the reduction was for calcium tungstate screens than for the rare earth
less (-15%). screens [ 5 ] . Thus, the contrast reduction effect is likely
to be less marked for calcium tungstate screens. In gen-
Table I. Theoretical and measured contrast at different tube eral, the magnitude of the effect would vary with the
potentials (see text for definition of parameters) screen and possibly also the tube potential. It has been
observed that for screens containing the same phosphor,
kVp HVL n C C c/cn the higher the screen speed, the greater was the depen-
(mm) (mm) ~ dence of speed on beam quality [ 5 ] . Hence, the magni-
tude of the contrast reduction effect may increase with
40 1.85 0.375 2.96 0.96 2.56 0.83 0.81 1.02 the speed of the screen.
60 2.99 0.232 3.04 0.61 2.24 0.45 0.47 0.96
80 3.70 0.187 2.99 0.49 2.06 0.34 0.33 1.03
100 4.70 0.147 2.97 0.38 2.11 0.27 0.25 1.08 References
120 5.30 0.131 3.04 0.35 2.09 0.24 0.23 1.04 1. REYNOLDS, J, SKUCAS, J and GORSKI, J, An evaluation
140 5.75 0.121 3.04 0.32 1.95 0.2 0.22 0.91 of screen-film speed characteristics, Radiology, 118, 711-713
(1976).

Vol. 69, No.822 565


2. VYBORNY, C J, METZ, C E, DOI, K and ROSSMAN, K,
Screen/film system speed: its dependence on X-ray energy,
Radiology, 125, 811-816 (1977).
3. VENEMA, H W, X-ray absorption, speed, and luminescent
efficiency of rare earth and other intensifying screens,
Radiology, 130, 765-771 (1979).
4. THUNTHY, K H and WEINBERG, R, Effect of kilovoltage
on the relative speed of rare earth screens, Dentomaxillofac.
Radiol.,15, 27-30 (1986).
566
G A Shah, G Hassam and D L Newman
5. KODERA, Y, DOI, K and CHAN, H P, Absolute speeds.of
screen-film systems and their absorbed-energy constants,
Radiology, 151, 229-236 (1984).
6. WOLBARST, A B, Physics of Radiology (Appleton & Lange,
Norwalk)(1993).
7. VYBORNY, C J, H and D curves of screen-film systems:
factors affecting their dependence on X-ray energy, Med.
Phys., 6, 39-44 (1979).
The British Journal of Radiology, June 1996