Wednesday Case of the Day
Category: Physics
Sara C. Gavenonis MD, Andrew D.A. Maidment PhD
Hospital of the University of Pennsylvania, Philadelphia, PA
History: 45 year-old female enrolled in an imaging research study.
A
B
C
1. There is an increase in
noise between A and
B. What radiographic
factor would most
directly account for
this difference?
2. There is a decrease in
subject contrast
between B and C.
What radiographic
factors could account
for this difference?
Question #1 Discussion: The x-ray tube current (mAs) directly
controls the noise seen in an x-ray or CT image.
A
B
Figure 2: These two images show the
effect of reducing the tube current
(mAs) on image quality.
A) This image was acquired at 28
kVp with a molybdenum x-ray
target and a rhodium filter. The
image was acquired at 48 mAs.
B) This image was acquired at the
same kVp, target and filter as
image A. However, the image was
acquired at 16 mAs, or one-third
the dose.
Digital breast tomosynthesis will be discussed in this exhibit.
Please note that as of 11-9-2010, digital breast
tomosynthesis is not FDA approved.
Question #1 Discussion: The noise in an image is
determined by the dose. The most direct determinant
of radiographic dose is the x-ray tube current or mAs.
28 kVp
8 mAs
0.13 mGy
28 kVp
70 mAs
1.13 mGy
28 kVp
220 mAs
3.39 mGy
Figure 3: A mammographic phantom is shown imaged at 3 different mAs values. As the
mAs (and hence dose) increases, the number of objects visible in the phantom increases
too. The typical diagnostic technique is 70 mAs.
Question #1 Discussion: Reduction of tube current will
adversely affect the ability to detect fine linear structures
such as fibrils, and micro-calcifications (arrow). In
general, large objects are not affected (double arrow)
Figure 4: A normal mAs (left) and low mAs (right) image of a calcification cluster.
Question #1 Discussion: This patient was enrolled in a research study in
which digital mammography and digital breast tomosynthesis exams
were obtained.
A
B
A) Digital mammography image.
B) Digital breast tomosynthesis
source projection image .
Please note that as of 11-9-2010, digital breast
tomosynthesis is not FDA approved.
Discussion: The source
projection images used to
create tomosynthesis image
reconstructions are acquired at
low dose. The total dose of a
tomosynthesis image set is
comparable to the dose of a
mammogram. Since 10-25
source images are used in the
reconstruction, each source
image is acquired at 1/10th to
1/25th the mammographic dose
Figure 5: A tomosynthesis image is
acquired as a series of low dose
projection images. Here 9 source
positions are shown.
Discussion: While each tomosynthesis projection image
is noisy, the tomosynthesis reconstruction process
reduces the noise in the reconstructed image
Figure 6: The tomosynthesis source projection image (left) is noisy, limiting the
detection of the calcifications. However, the tomosynthesis reconstruction combines
the information from many source projection images. Therefore, the resulting
reconstruction (right) has low noise and more calcifications are visible.
Question #2 Discussion: The contrast of an image is
determined by the kVp and target/filter combination.
B
C
Figure 7: These two images
show the effect of changing
the target material, filtration,
and kVp on image quality.
B) This image was acquired at
28 kVp with a molybdenum
x-ray target and a 0.025 mm
thick rhodium filter.
B) This image was acquired at
49 kVp with a rhodium x-ray
target, a 0.025 mm thick
rhodium filter and a 0.3 mm
thick copper filter.
The dose to the breast in
these two images is
approximately the same.
Question #2 Discussion: Changing the kVp or target/filter
combination will alter the contrast of an image.
24 kVp
185 mAs
1.2 mGy
28 kVp
71 mAs
1.2 mGy
34 kVp
30 mAs
1.2 mGy
Figure 8: A mammographic phantom is shown imaged at 3 different kVp values. As the kVp
increases, the mAs was decreased to keep dose constant. The optimal image is at 28 kVp. The
24 kVp image is noisy because the number of x-rays transmitted through the phantom is
reduced. The 34 kVp image is noisy because the x-ray subject contrast is lower; to get sufficient
display contrast, the window width was reduced which amplifies the appearance of the noise.
Discussion: High energy (high kVp) x-ray images of the breast are
being used in clinical trials for contrast-enhanced digital
mammography and contrast-enhanced digital breast
tomosynthesis.
HE
HE
10000
1800000
1600000
1000
1400000
K-edge
1000000
Iodine
800000
10
600000
Breast
HE
spectrum
1
Pre contrast
1200000
100
400000
Post contrast
200000
0.1
0
0
10
20 keV 30
40
50
Time
Figure 9: The high energies emphasize the attenuation of iodine, increasing the contrast of
enhancing lesions. However, high energy images do not preserve breast tissue contrast.
# Photons
Mass Attenuation [cm2/g]
Iodine
contrast
Summary: The x-ray spectrum determines the contrast
and noise seen in radiographic and tomographic
images, as well as the dose to the patient. The
x-ray spectrum is controlled by the tube current
(mAs), tube voltage (kVp) and target material, and
the filtration added to the x-ray beam.
References:
1)
JT Bushberg, et al. The Essential Physics of Medical Imaging, 2nd Edition.
2)
LT Niklason, et al. Digital tomosynthesis in breast imaging. Radiology 1997;
205:399-406.
3)
SC Chen, et al. Initial clinical experience with contrast-enhanced digital breast
tomosynthesis. Academic Radiology 2007; 14: 229-238.
Acknowledgements: The clinical images shown in this exhibit were processed and
reconstructed using the AdaraGPU™ and Briona™ software from Real Time Tomography,
LLC. (Villanova, PA) and have been provided courtesy of Susan Ng.