Download How does magnification affect image quality and patient dose in

How does magnification affect image quality and patient dose
in digital subtraction angiography?
Nikolaos A Gkanatsios*a, Walter Hudab, Keith R Petersc
a
b
c
Medical Physics (H-539), Baylor University Medical Center
3500 Gaston Avenue, Dallas, TX 75246
Department of Radiology, SUNY Upstate Medical University, Syracuse, NY 13210
Department of Radiology, University of Florida, PO Box 100374, Gainesville, FL 32610
ABSTRACT
Digital subtraction angiography (DSA) images were obtained of a phantom containing 1 mm diameter vessels. The iodine
concentrations ranged from 5 to 50 mg/cc, which permitted the determination of the detection threshold concentration of
iodine. The source to image receptor distance was 105 cm, and image magnification was varied between 1.15 and 2.0. One
experiment was performed at an input exposure of 1 µGy per frame, and a second experiment was performed at 4 µGy per
frame. Surface (skin) doses were measured using an ionization chamber, and the corresponding values of energy imparted
were determined from the exposure area product. Increasing the radiation exposure by a factor of four reduced the threshold
iodine concentration by ~30%. The average detection threshold iodine concentration for a magnification of 1.15 was 13.6
mg/cc. Detection performance improved linearly with magnification, with an average value of 7.4 mg/cc at a magnification
of 2.0. Values of energy imparted were essentially independent of image magnification, whereas surface doses increased by
a factor of four when image magnification was increased from 1.15 to 2.0. Magnification offers improved imaging
performance at no additional patient risk provided that surface doses do not exceed the dose threshold for deterministic
effects such as skin burns and epilation.
Keywords: magnification, dosimetry, image quality, neuroradiology, digital subtraction angiography.
1. INTRODUCTION
Diagnostic and therapeutic interventional neuroradiologic procedures involve imaging of catheter manipulation and vascular
anomalies of the brain.1,2 These interventional neuroradiologic procedures often involve long fluoroscopic exposure times
and the acquisition of a large number of radiographic images.3 As a result, there is a possibility of induction of deterministic
radiation effects such as skin erythema and epilation, as well as producing a relatively high stochastic risk.4,5
Magnification and its effects on image quality have been studied in conventional radiography.6,7,8 In general, magnification
improves visibility of small, low contrast objects. As the magnification increases, the effective noise in the image detector is
reduced improving the signal-to-noise ratio, and visibility of small structures improves.9 Scatter radiation is also reduced
with increased magnification, which improved contrast detectability.10
Geometric magnification can also improve detection of small vessels during interventional neuroradiologic procedures using
digital subtraction angiography (DSA). Modification of geometric magnification will impact on image quality and patient
doses during interventional neuroradiologic procedures. Quantifying the effect of magnification on image quality and patient
doses will help optimize these types of radiologic examinations. In this study, we investigated the relationship between
geometric magnification, image quality and the corresponding patient doses.
*
Correspondence: Email: [email protected]; WWW: nersp.nerdc.ufl.edu/~nikos; Telephone: 214 820 7143; Fax: 720 221 4942.
326
2. METHOD
2.1. Phantom
Figure 1 illustrates the phantom used to simulate 1.0 mm diameter vessels for the purpose of evaluating image quality in
neuroradiology. The phantom consists of stacked acrylic blocks with dimensions of 30 cm × 30 cm × 1.3 cm. An insert
holder made of acrylic with a thickness of 1.3 cm is positioned at the center of the phantom to accommodate a vessel insert.
A blank and a vessel insert measuring 30 cm × 9.0 cm × 1.3 cm were made out of acrylic. The blank insert was used to
acquire mask images during digital subtraction angiography.
The vessel insert had thirty cylindrical vessels 1.0 mm in diameter and 35 mm in length drilled along its midplane at intervals
of 8.0 mm apart. The total phantom thickness was 16.5 cm of acrylic, equivalent to 20 cm of water. The vessels on the
acrylic insert were filled with iodinated contrast with iodine concentrations ranging from 50 mg/cc to 5.0 mg/cc. The iodine
concentration in each vessel was 92% of the previous concentration.
2.2. DSA Acquisitions
The experimental setup used an x-ray source to image receptor distance (SID) of 105 cm (maximum SID). The 23 cm
diameter image intensifier mode was used for all image acquisitions. DSA acquisitions were performed using the 0.6 mm
focal spot size and 70 kVp. A 10×5-60 ionization chamber of a MDH 1015C exposure meter was attached to the surface of
the image intensifier behind the grid to record the input exposure to the image receptor.
The effect of image magnification was studied by varying the distance of the phantom relative to the image intensifier
entrance surface. The position of the phantom was varied to achieve a range of geometric object magnification from 1.15 to
2.0. During all digital subtraction acquisitions, a mask of a single frame was acquired at 3 frames/sec acquisition rate using
the blank phantom insert. The blank insert was then replaced by the vessel insert and twenty additional frames were acquired
with no frame integration at the same acquisition rate of 3 frames/sec. The image intensifier input exposure was first set to 1
µGy/frame and the experiment was repeated at 4 µGy/frame. The mAs was adjusted accordingly to produce a mean pixel
value between 2000 and 2100 as shown in Table 1.
TABLE 1. Imaging Techniques During Geometric Object Magnification Experiments
Mag
1.15
1.20
1.40
1.60
1.80
2.00
1 µGy/frame
Video
mAs
Level
4.40
2100
4.60
2068
4.80
2008
5.00
2036
5.40
2000
5.80
2072
4 µGy/frame
Video
mAs
Level
16.3
2100
17.0
2018
17.8
2000
18.3
2010
18.8
2000
19.3
2000
2.3. Dosimetry
The exposure to the phantom was measured at the x-ray beam entrance plane using a 10×5-6 ionization chamber of an MDH
1015C. The measured exposure included backscatter radiation coming from the acrylic phantom. The entrance exposure per
frame was obtained from the integral exposure measured for each image acquisition sequence divided by the number of
acquired frames. A conversion factor of 2.58×10-4 C/kg (1R) corresponding to an absorbed dose of 9.3 mGy for muscle
tissue was used to convert the entrance exposure to surface dose.
The energy imparted was computed using the method described by Gkanatsios and Huda11 for the applied tube voltage and
corresponding half-value layer. The exposure was obtained from direct exposure measurements during each image
acquisition sequence and included backscatter. The backscatter radiation fraction was measured for the corresponding
phantom and applied tube voltages and was subtracted from the measured exposures to obtain the free-in-air exposure. The
exposure area at the beam entrance plane of the phantom was computed from geometry assuming that the beam area at the
image intensifier plane was a circle with a diameter of 23 cm.
327
2.4. Image Quality Evaluation
The threshold iodine contrast concentration was determined from the subtracted images of the vessel insert using a single
(blinded) observer. A five-point scale was used to determine the visibility of any vessel containing iodine contrast ranging
from 1 (vessel invisible) to 5 (vessel perfectly visible). A rank of 3 corresponded to a vessel between these two extremes,
where the observer was 50% confident of seeing the vessel. The observer was trained on the criteria of scoring each iodine
concentration using ten subtracted images similar to those asked to evaluate, and requested to be consistent in the application
of the scoring scheme. All the images were then presented in a random order to the trained observer who selected the
concentration that corresponded to the rank value of 3. The window and level of the displayed images were adjusted to
optimize signal detection during each DSA image acquisition, so that the resulting images would not be contrast limited.
3. RESULTS AND DISCUSSION
3.1. Imaging Performance.
Figure 2 shows the threshold iodine concentration as a function of geometric object magnification. The precision of the
threshold contrast determination was ~12% when the detection threshold was ~8 gm/cc. The difference in threshold
concentration between the 1 µGy/frame and 4 µGy/frame input exposure was ~30% across the magnification range shown in
Figure 2.
The average detection threshold iodine concentration for a magnification of 1.15 was 13.6 mg/cc. Detection performance
improved linearly with magnification, with an average value of 7.4 mg/cc at a magnification of 2.0. The relationship between
magnification and threshold iodine contrast was approximately linear with the coefficient of determination of the least
squares fit (r2) equal with 0.53 for the 1 µGy/frame an 0.92 for the 4 µGy/frame. An increase by a factor of two to geometric
magnification decreased threshold concentration by approximately one half.
Image quality is very critical in interventional neuroradiologic procedures. The ability to visualize small and low contrast
objects is of paramount importance, where neurovascular instruments may be ~200 µm and vessel sizes may be ~100 µm.
Any dose reduction strategy must always ensure that image quality is not compromised and patients do not suffer any adverse
clinical consequences as a result of inadequate visualization of catheters or vasculature.
3.2. Patient Dose.
Figure 3 shows the surface dose and energy imparted as a function of geometric object magnification at 1 µGy/frame and 4
µGy/frame. The measured precision of the surface dose measurements was ~0.7%. Surface dose increased as the square of
increase in geometric magnification due to its inverse proportionality to the square of the source-to-surface distance. On the
other hand, energy imparted remained almost constant with geometric magnification. The 10% increase in energy imparted
across the magnification range shown in Figure 3 was caused by the increased in imaging techniques to maintain a constant
video level. The increase in radiographic techniques was caused by the reduced scatter reaching the image intensifier by
moving the phantom farther away from the image intensifier to increase magnification.
The surface dose predicts the possibility of inducing deterministic injuries from high dose interventional radiologic
procedures (i.e., cardiac catheterization, abdominal interventional or neurointerventional procedures). Deterministic injuries
associated with interventional neuroradiologic procedures primarily consist of injuries induced to the skin of the patient,
including skin erythema and epilation. Below the threshold dose for the induction of deterministic effects, the patient risk is
primarily from the stochastic processes of carcinogenesis and the induction of genetic effects.12 The energy imparted to the
head may be taken to be proportional to the stochastic radiation risk associated with this type of x-ray examination for any
given patient.
In neurointerventional radiology, magnification is often used as a tool to visualize small vasculature. Care should be taken,
however, when magnification is used, since the surface dose to the patient increases significantly with magnification. Energy
imparted, on the other hand, is independent of magnification as distance from the x-ray source and area of exposure decrease
equally as magnification is employed. The results of this study clearly demonstrate that magnification offers improved
imaging performance at no additional patient risk provided that surface doses do not exceed the dose threshold for
deterministic effects.
328
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
F. Vinuela, Interventional Neuroradiology, Raven Press, New York, 1992.
A. G. Osborn, Diagnostic Cerebral Angiography, Second Edition, Lippincott, Philadelphia, 1999.
N. A. Gkanatsios, W. Huda, K. R. Peters, and J. A. Freeman, “Evaluation of an on-line patient exposure meter in
neuroradiology,” Radiology 203, pp. 837-842, 1997.
W. Huda, and K. R. Peters, “Radiation-induced temporary epilation after a neuroradiologically guided
catheterization procedure,” Radiology 193, pp. 642-644, 1994.
L. K. Wagner, P. J. Eifel, and R. A. Geise, “Potential biological effects following high x-ray dose interventional
procedures,” J of Vasc Int Rad 5, pp. 71-84, 1994.
K. Doi, and K. Rossmann, “The effect of radiographic magnification on blood vessel imaging with various screenfilm systems,” Medical Physics 1, pp. 257-261, 1974.
L. K. Wagner, G. Cohen, W. H. Wong, and S. R. Amtey, “Resolution in radiographic magnification,” Medical
Physics 8, pp. 382-384, 1981.
J. M. Sandrik, and R. F. Wagner, “Absolute measures of physical image quality: measurement and application to
radiographic magnification,” Medical Physics 9, pp. 540-549, 1982.
K. Doi, and K. Imhof, “Noise reduction by radiographic magnification,” Radiology 122, pp. 479-487, 1977.
T. Sandor, and P. Natt, “Effect of radiographic magnification on image contrast of blood vessels,” American Journal
of Roentgenology 134, pp. 159-192, 1980.
N. A. Gkanatsios, and W. Huda, “Computation of energy imparted in diagnostic radiology,” Medical Physics 24, pp.
571-579, 1997.
International Commission on Radiological Protection (ICRP) Publication 60. Pergamon Press (Oxford) 1991.
Digital Subtraction Angiography Phantom
Interchangeable
FIGURE 1. Schematic diagram of the acrylic phantom with the vessel and blank inserts used to simulate small vessels for the purpose of
evaluating image quality in neuroradiology.
329
Iodine Concentration (mg/cc)
24
24
18
18
1 µG/frame
12
12
6
6
4 µG/frame
0
0
1.0
1.2
1.4
1.6
1.8
2.0
Geometric Object Magnification
FIGURE 2: Threshold iodine concentration as a function of geometric object magnification at 70 kVp.
1.6
1.6
1.2
1.2
0.8
0.8
0.4
0.4
Energy Imparted
Surface Dose
0.0
Energy Imparted (mJ/frame)
Surface Dose (mGy/frame)
1 µG/frame
0.0
1
1.3
1.6
1.9
2.2
Geometric Object Magnification
FIGURE 3: Surface dose and energy imparted as a function of image intensifier input exposure at 70 kVp.
330