Improved fracture detection using the mammographic film-screen combination
Y
Faridah,MBBS, MRad, BJJ Abdullah, MBBS,
FRCR, KH Ng, PhD, MIPEM, DABMP
Department of Biomedical Imaging
(Radiology), Faculty of Medicine, University Malaya, Kuala
Lumpur, Malaysia
ABSTRACT Aim:
The single emulsion or single screen system is usually reserved
for mammography since its use in general radiography is limited.
The purpose of this study is to compare the mammographic film-screen
combination (MFC) and the standard film-screen combination
(SFC) in terms of fracture and soft tissue injuries detection.
Patients, methods and materials: In this
prospective study, 41 patients from Accident and Emergency
suspected of having injury in the hands, wrists, ankles and
feet regions were radiographed using both MFC and SFC. These
were compared in terms of image quality, presence of fractures
and soft tissue injuries. The two different film-screen combinations
were also compared in terms of detection of bony fragments,
film characteristics such as film speed, contrast and spatial
resolution, dose and cost.
Results: The MFC gives statistically better
image quality compared to SFC. In 10% of patients, fractures
were detected only in the MFC, which also detects tiny bone
fragments that may not be resolved by the SFC. The spatial
resolution of the MFC is greater than the SFC. The film speed
and contrast of the MFC are lower than that of the SFC. The
doses of MFC were higher compared to SFC.
Conclusions: The MFC detects fractures better
compared with SFC. However, the entrance skin dose for the
mammographic film-screen combination was about 35% to 55%
higher than the standard film-screen combination.
Keywords: Fracture, detection, film-screen combination,
mammography
INTRODUCTION Plain radiography
is an important diagnostic tool in the Accident and Emergency
Department (A&E). The effectiveness of the radiograph
as a diagnostic tool is firstly dependent on good radiological
quality and secondly on minimization of oversight in fracture
detection. On admission, 23% of fractures are overlooked on
initial radiographs by radiologists resulting in mismanagement
of patients. Some of these fractures were missed because of
technically inadequate X-rays while others with adequate X-rays
had fractures that could not be identified on admission films
[1]. Clearly, a higher quality radiograph
could be instrumental in achieving higher rate of detection
of fractures and soft tissue injuries and hence resulting
in better management.
The purpose of this study is to compare both the detection
of fractures and soft tissue injuries using a mammographic
film-screen combination (MFC) and the standard film-screen
combination (SFC). The abilities of these two film-screen
systems in detecting fractured fragments in vivo are also
analyzed. The characteristics of these two film-screen systems
i.e. contrast, speed, spatial resolution, entrance skin dose
as well as cost were compared.
PATIENTS AND METHODOLOGY
Patient selection
A prospective study was carried out on patients referred from
A&E, for suspected bony or soft tissue injuries involving
the wrist, hand, ankle and foot. A total of 41 patients were
randomly selected. Informed consent was obtained from patients
or their guardian. The patients’ age ranged from 17
years to 66 years with a mean of 43 years. Majority of the
patients were male (80.5%). This study was approved by the
hospital ethics committee.
Image quality and detection of fractures and soft
tissue injuries
A pair of radiographs using a standard radiographic combination
(Kodak T-Mat G film, Kodak Lanex Regular Screen) and a mammographic
film-screen combination (Fuji UM-MA HC film, Kodak Min–R
Screen) was taken. For each area of interest, two identical
tangential views were taken in both the mammographic and conventional
film-screen combinations (Table 1).
The focus-film distance (FFD) was kept at 100 cm while the
film size was 24 cm x 30 cm. The radiographic factors for
MFC were predetermined using a phantom while exposure factors
of the SFC were those currently being used in our department
(Table 1). Both types of film were then processed using the
same 90s processor (Kodak M6B RPX-Omat) at a developing temperature
of 32.5 deg C.
Two radiologists assessed the films independently by completing
questionnaires on image quality, and detection of fractures
and soft tissue injuries (Table 2). For the purpose of quality,
“gold standard” images deemed to display the best
image quality from each sets of SFC and MFC films were chosen
as the reference [2].
[View this table] |
Table 2 Questionnaire for (a)
independent rating of image quality; (b) detection
of fractures and soft tissue injuries |
|
The radiographs from SFC were assessed and rated while the
radiographs from MFC were assessed two weeks later. Subsequently,
another questionnaire was completed to assess the two sets
of film in terms of imaging quality and detection of fractures
(Table 3).
[View this table] |
Table 3 Questionnaire for comparable
rating of (a) image quality; (b) detection of fractures
and soft tissue injuries |
|
In vivo detection of bone fragments
A piece of cadaveric metacarpal bone was intentionally broken
into fragments of different sizes to simulate bony fractures.
These bone fragments were measured in length and width (mm)
and suspended in a wax mould to approximate the soft tissue
density of the extremities. This phantom was exposed using
both MFC and SFC. The number and size of the fragments visualized
in each type of film were noted and compared.
Assessment of film characteristics
The film characteristics i.e. the characteristic curve (H&D
curve), speed, contrast, spatial resolution and dose of each
film-screen combination were assessed as below.
Characteristic curve, speed and contrast – using a sensitometer,
an optical density versus step number graph were obtained
for both single and double film-screen combination. From this
the speed, contrast and base plus fog level were obtained.
Using a resolution metal bar test object manufactured by Nuclear
Associates New York, Nr 86633, the two film-screen combinations
were exposed. The spatial resolution for each system was then
derived.
Using a dose meter, the radiation dose in terms of mGy/mAs
was calculated for the different exposures. The entrance skin
dose in mGy for each region was then obtained.
RESULTS
Image quality and detection of fractures and soft
tissue injuries
The scores were tabulated and statistically analyzed using
the Wilcoxon Signed Ranks Test [StatView software 5.0 (SAS
Institute Inc. North Carolina)]. Hand and wrist radiographs
were performed in 25 patients, ankle radiographs in 12 patients
and foot radiograph in four patients. The hands and wrists
were rated together as the size and thickness of these two
regions are almost similar. The image quality of the MFC radiographs
for the hands and wrists were statistically better (p<0.05)
compared with the SFC radiographs (Figures 1a and 1b).
[View this figure] |
Figure 1 Image quality was better
on the (a) MFC radiograph compared with (b) SFC
radiograph. Soft tissue and outline of carpal bones
are better depicted in MFC radiograph as well. Note
that break in the cortex at distal radius with overlapping
of fracture outline (arrow) is better appreciated
in the MFC radiograph |
|
For the foot and ankle regions however, the difference in
image quality of radiographs taken using the two different
systems was not statistically significant.
In the 41 patients seen, fractures were detected in 20 patients,
giving a detection rate of 48.8%. Both fractures and soft
tissue injuries were statistically better detected (p<0.05)
on MFC compared with SFC for the hand and wrist regions (Figures
2a and 2b). Again, the ankle and foot regions showed no statistical
significance in depicting fractures and soft tissue injuries
between the two film-screen combinations. Interestingly, in
the 20 patients with fractures, two of these fractures were
seen on the MFC but not on the SFC resulting in a 10% increase
in the detection of fracture with MFC (Figures 3a and 3b).
[View this figure] |
Figure 2 Fracture at distal phalanx
of the ring finger (arrow) was better outlined due
to better spatial resolution in (a) MFC compared
to (b) SFC |
|
[View this figure] |
Figure 3 Fracture at distal posterior
tibia (arrow) was only seen on MFC image (a) but
was missed on SFC image (b) |
|
In-vivo detection of bone fragments
The MFC was able to resolve up to the second smallest fragment
i.e. 0.7 mm x 1.0 mm whereas the SFC could only display up
to the third smallest fragment measuring 1.0 mm x 2.0 mm.
Furthermore the specks of bone in the centre were better appreciated
in numbers and delineation on the MFC radiograph (Figures
4a and 4b).
[View this figure] |
Figure 4 Images of bones intentionally
broken then suspended in a phantom. Specks of bone
fragments in the centre are better resolved in the
MFC image (a) compared to the SFC image (b) |
|
Film characteristics
MFC and SFC was compared in terms of film curve, film contrast,
film speed, spatial resolution, entrance skin dose and cost.
The base plus fog levels of these two films were almost similar
i.e. 0.27 for MFC and 0.25 for SFC. The curve of the single
emulsion film is steeper than the double emulsion film. The
contrast index of MFC is 1.53 while the contrast index for
SFC is 1.88. These are comparable to the steepness of the
different curves for the different film-screen combinations
seen. The contrast of MFC is comparable to SFC up to the density
of 3.0. At higher densities, the contrast of SFC is higher.
The speed indices for MFC and SFC are 1.20 and 2.44 respectively.
Using the resolution bar test object, the spatial resolutions
of both systems were attained. The spatial resolution of MFC
is 10 lp/mm whereas the resolution of SFC is 6 lp/mm.
The entrance skin doses for MFC were higher compared to
SFC in all the regions radiographed. There was a 34% to 40%
increase in entrance surface dose in the hands and wrists
region on using MFC. In the foot, the increase in the entrance
skin dose was approximately 34%. The ankle documented the
highest increase in the entrance skin dose of 54% to 57% on
using MFC compared to SFC (Figure 5).
[View this figure] |
Figure 5 Bar chart of entrance
skin doses by region of interest. There was a 34%
to 40% increase in entrance surface dose in the
hands and wrists region on using MFC |
|
The cost of a box of 100 Fuji UM-MA HC films (single-emulsion)
measuring 24 cm x 30 cm is approximately twice higher compared
with the cost of a box of 100 Kodak T-Mat G films (double-emulsion)
of similar dimension.
DISCUSSION
In an outpatient setting, it had been found that the most‘missed’diagnosis
is a fracture [3]. Furthermore, incorrect
diagnoses were seen most frequently in the more commonly injured
anatomical sites – ankles, wrist, foot, elbow and hand
[4]. For example, scaphoid fractures were
reported to represent 2% to 7% of all fractures and over 70%
of all hand fractures presenting to accident and emergency
(A&E) departments [5]. Unfortunately,
conventional X-rays miss up to 2% of these fractures on the
first presentation [6]. This is further
compounded by the fact that emergency physicians are the first-line
radiograph interpreter in an A&E setting. Rates of disagreement
between emergency physicians and radiologists in the interpretation
of skeletal radiographs have been documented to range from
9% to 10% [7]. A change of treatment was
required for 1% to 3% of these patients [8].
Although both MRI and bone scintigraphy have been shown to
display fractures earlier and with greater accuracy than conventional
radiography [9], these imaging modalities
are expensive and are not easily available.
Why fractures are misdiagnosed is multifactorial. Attributions
have been made to poor clinical inspection or failure on the
part of physician to consider certain clinical entities [10],
failure to request for appropriate radiographic views [11]
or failure on part of clinician to recognize a fracture on
the radiograph [12]. The quality of the
radiographic image also plays a part in the misdiagnosis of
fractures. Review of the literature shows there are no references
made to the type of the film-screen combination used in these
instances where fractures were missed. Since the double film-screen
combination is the most conventional system, it is fair to
assume that in these cases, a SFC was employed. SFC is commercially
available, has a wide exposure latitude, is easy to process
and the dose to patient is reduced due to less X-ray photons
used. However, SFC due to its front and back luminescent screens,
give rise to lack of image sharpness. MFC, on the other hand
due to its high-resolution capabilities, would make a reasonable
alternative.
In comparing SFC with MFC, it was found that MFC generally
gave a better image quality. This is attributed to the higher
spatial resolution of MFC resulting in clearer outlines of
structure. The absence of the front screen in an MFC system
decreased the parallax effect, which is responsible for producing
a shift in the image on two sides of film [13].
Fractures and soft tissue injuries were also better seen on
MFC, where 10% of these fractures were seen only on MFC. In
these two cases, the doctors attending to these patients in
the A&E were informed of the fractures and proper treatment
was instituted. These patients under normal circumstances
would have been discharged, only to subsequently return with
increased morbidity. These findings have been confirmed by
the in vivo study where MFC detected small fragments of bone
and specks of calcification better than SFC. A study by Oestmann
et al. revealed similar results when film geometry was similar
i.e. no magnification [14]. This indicated
that in a clinical setting, tiny bone fragments (from an avulsion
fracture perhaps) would be better detected if MFC rather than
SFC was used.
The shape of the H&D curve is usually independent of
the screen used and is determined only by the characteristics
of the film and the processing condition [13]. Although the
mammographic film is usually processed using a dedicated 2.5min
processor in mammography, the film may be developed using
the standard 90s processor. The base plus fog levels remained
unchanged; while the speed and contrast indices were increased
[15]. Film contrast is higher with SFC
compared with MFC especially with densities higher than 3.0.
This renders MFC with a narrow exposure latitude and thus
there is not much flexibility in the exposures that can be
used to form an image of optimal density. This is a major
drawback, as accurate exposure factors need to be ensured
before exposing the patient. This is difficult to determine
as patients come in different sizes and shapes. This was not
noticeable in the hand and wrist radiographs but proved detrimental
in imaging of thicker parts such as the foot and ankles. Film
speed is higher for SFC compared with MFC. This is because
removal of the front screen in MFC decreased the number of
light phosphor being emitted from the screen, but this decreased
the crossover exposure effect commonly seen with SFC, significantly
increasing its resolution [13, 16].
A major drawback is the increased radiation dose with MFC.
This is because of the decrease in speed, which therefore
increased the exposure time. There is also a concomitant increase
in kVp for MFC (about 20 kV higher). The other drawback of
the MFC system is its cost. Although it is slightly more expensive
than SFC [15], the huge number of radiographs
exposed in a large hospital everyday may incur huge costs.
There would also be added costs in having to purchase new
cassettes for this endeavour. In addition, the types of single
film available in a country such as ours, is limited.
The paediatric patients were excluded from the study, as
the exposure factors required for imaging of paediatric group
would be different. It would be interesting to document whether
MFC could show fractures better in this group. The study had
also been limited to the extremities of upper and lower limbs.
Further work with the elbow and knee regions could be carried
out. This could prove beneficial especially in the knee as
a number of fractures and injuries in this region are difficult
to diagnose and remain undetected on SFC. The ability of MFC
to display fractures and soft tissue injuries could be compared
with other imaging modalities that have been shown to be superior
to plain radiographs such as bone scan and magnetic resonance
imaging.
CONCLUSION
MFC gives better visualization in terms of image quality
and fracture detection in the hand and wrist regions, compared
with SFC. It also gives better detection of tiny bone fragments
that may not be resolved by SFC. The film speed and contrast
of MFC are lower than that of SFC. The spatial resolution
of MFC is greater than SFC.
The radiation dose to the patient is however increased with
MFC. Although radiation dose in imaging of the extremities
is limited [15], patients confirmed to have fractures would
need repeated radiographs and hence the radiation dose would
add up considerably. Nevertheless, we believe that MFC would
contribute significantly in reducing misdiagnosis of fractures
as well as increasing the detection of avulsion injuries.
References
-
Born CT, Ross SE, Iannacone WM, et al. Delayed identification of skeletal injury in multisystem trauma: the 'missed' fracture. J Trauma 1989;29(12):1643-6.
[Medline]
-
Britton CA, Gabriele OF, Chang TS, et al. Subjective quality assessment of computed radiography hand images. J Digit Imaging 1996;9(1):21-4.
[Medline]
-
Johansson H, Raf L. A compilation of "diagnostic errors" in Swedish health care. Missed diagnosis is most often a fracture. Lakartidningen 1997;94(43):3848-50.
[Medline]
-
Thomas HG, Mason AC, Smith RM, et al. Value of radiographic audit in an accident service department. Injury 1992;23:47-50.
-
Grover R. Clinical assessment of scaphoid injuries and the detection of fractures. J Hand Surg Br 1996;21(3):341-3.
[Medline]
-
Parvizi J, Wayman J, Kelly P, et al. Combining the clinical signs improves diagnosis of scaphoid fractures. A prospective study with follow-up. J Hand Surg Br 1998;23(3):324-7.
[Medline]
-
Robinson PJ, Wilson D, Coral A, et al. Variation between experienced observers in the interpretation of accident and emergency radiographs. Br J Radiol 1999;72(856):323-30.
[Medline]
-
Lufkin KC, Smith SW, Matticks CA, et al. Radiologists' review of radiographs interpreted confidently by emergency physicians infrequently leads to changes in patient management. Ann Emerg Med 1998;31(2):202-7.
[Medline]
-
Pillai A, Jain M. Management of clinical fractures of the scaphoid: results of an audit and literature review. Eur J Emerg Med 2005;12(2):47-51.
[Medline]
-
Moore MN. Orthopedic pitfalls in emergency medicine. South Med J 1988;81(3):371-8.
[Medline]
-
Zeitoun F, Frot B, Sterin P, et al. Views necessary for the traumatic wrist. Ann Radiol (Paris) 1995;38(5):255-65.
[Medline]
-
Saab M, Stuart J, Randall P, et al. X-ray reporting in accident and emergency departments--reducing errors. Eur J Emerg Med 1997;4(4):213-6.
[Medline]
-
Higashida Y, Frank PH, Doi K. High-speed, single-screen/single-emulsion film systems: basic imaging properties and preliminary clinical applications. Radiology 1983;149(2):571-7.
[Medline]
-
Oestmann JW, Kopans DB, Linetsky L, et al. Comparison of two screen-film combinations in contact and magnification mammography: detectability of microcalcifications. Radiology 1988;168(3):657-9.
[Medline]
-
Abdullah BJ, Kaur H, Ng KH. An in vitro study comparing two different film-screen combinations in the detection of impacted fish bones. Br J Radiol 1998;71(849):930-3.
[Medline]
-
Haus AG. The AAPM/RSNA physics tutorial for residents. Measures of screen-film performance. Radiographics 1996;16(5):1165-81.
[Medline]
Received 12 April 2005; received
in revised form 6 June 2005; accepted for publication
10 June 2005
Correspondence: Department of Biomedical Imaging (Radiology), University of Malaya Medical Centre, 59100 Kuala Lumpur, Malaysia.. Tel.: 603-79502069; Fax: 603-79581973; E-mail: yangf@um.edu.my (Yang Faridah A. Aziz).
Please cite as: Faridah Y, Abdullah BJJ,
Ng KH,
Improved fracture detection using the mammographic film-screen combination, Biomed Imaging Interv J 2005; 1(1):e3
<URL: http://www.biij.org/2005/1/e3/>
|
Leave a comment