Академический Документы
Профессиональный Документы
Культура Документы
Amato et al.
Relationship Between Contrast Media and Organ Dose in CT
T
he quantification and reduction of dose-length product and, hence, the overall
the absorbed radiation dose during effective dose both increase in a linear fash-
CT examinations is a key problem ion along with the number of scans.
when optimizing scan protocols, Radiation dose increases in iodine-charged
especially with the development of high-res- tissues and organs have been studied and pro-
olution MDCT scanners and when multiple- posed as a conceptual basis for contrast-en-
phase protocols are adopted, because evi- hanced radiotherapy. Such a technique was
Keywords: CT, iodinated contrast media, radiation dence of a potential hazard due to increments introduced many years ago, and several theo-
absorbed dose in the radiation dose delivered to the popula- retic and experimental studies have been made
tion as a result of the widespread use of CT is by means of Monte Carlo simulations, mea-
DOI:10.2214/AJR.12.8958
increasing [1–3]. In clinical practice, the ad- surements on phantoms, and in vitro or in vivo
Received March 25, 2012; accepted after revision ministration of iodinated contrast medium is experiments [5–11].
June 7, 2012. widely used to improve image quality and di- Both theoretic and phantom studies gener-
1
agnostic sensitivity. ally rely on the assumption of homogeneous
All authors: Section of Radiological Sciences,
The most widely recognized estimator of distribution of the high-Z element within the
Department of Biomedical Sciences and Morphologic
and Functional Imaging, University of Messina, Via the stochastic risk of radiation-induced can- target tissue and were aimed at determining
Consolare Valeria 1, Messina I-98125, Italy. Address cer inherent in CT examinations is the effec- dose enhancement factors in various irradia-
correspondence to E. Amato (eamato@unime.it). tive dose, evaluated from dose measurements tion geometries and regimens. Some micro-
and standardized conversion factors, which dosimetric studies about contrast-enhanced
Supplemental Data
Available online at www.ajronline.org. are calculated by using anthropomorphic radiotherapy have been conducted, and, con-
phantoms. The effective dose has been found cerning the x-ray energy spectrum emitted by
AJR 2013; 200:1288–1293 to have a good correlation with dosimetric the tube, they found that radiation quality (mean
0361–803X/13/2006–1288
quantities, such as the CT dose index and the lineal energy) effects introduced by the con-
dose-length product [4]. Consequently, when trast media are small compared with the dose-
© American Roentgen Ray Society multiple-phase CT studies are performed, the enhancement effect [10]. Furthermore, some
investigators found an increase in the frequen- trast medium during CT to be calculated. Here, we who wishes to apply the proposed method for the
cy of cellular abnormalities in patients who summarize the principles of the approach adopted quantification of the iodine concentration and the
underwent contrast-enhanced radiographic ex- and the results obtained in the previous study. Hounsfield unit increment, can measure γ accord-
aminations, and these results were supported We used a Monte Carlo simulation of an an- ing to the method described previously [25]. For
by radiobiologic experiments [12–22]. thropomorphic neck and abdomen phantom ex- our scanner operating at 120 kV, we found a value
Concerning the quantification of the ra- posed to a simplified model of a CT x-ray source of γ = 2.89 ⋅ 104 [24].
Downloaded from www.ajronline.org by 114.125.93.254 on 11/18/19 from IP address 114.125.93.254. Copyright ARRS. For personal use only; all rights reserved
diation dose enhancement in organs impreg- to obtain a relationship between the iodine mass Thus, we were able to find a relationship be-
nated by a contrast medium in contrast-en- accumulated by each organ during the scan and tween the Hounsfield unit increment and the rela-
hanced CT, the aforementioned CT dose the corresponding relative radiation dose incre- tive radiation dose increase for the organs studied:
index or dose-length product evaluations do ment (i.e., the difference between doses in the
∆D a ρ∆HU b
not account for the increase in organ dose presence and absence of contrast medium divided (4).
D ρTγ + ∆HU
due to the contrast agent, which accumulates by the dose in the absence of contrast medium):
at different concentrations in the acquisition ∆D = aφb (1), Because a, b, and ρT are constant characteristic
phases [23]. Indeed, the radiation dose in- D I of the tissue and γ has a constant value for a fixed
crease in an organ or tissue is dependent on where D is the dose, a and b are parameters that x-ray energy spectrum, Equation 4 allows the rel-
the quantity of iodine taken up at the time are characteristic of each organ, resulting from the ative dose increase from the increment in Houns-
of the scan; on anatomic factors such as the fit of Monte Carlo data with the analytical form of field units to be calculated.
organ shape, volume, and position inside the Equation 1, and the dimensionless quantity φI =
body; and finally, on the x-ray energy spec- mI / (mI + mT) is the iodine mass fraction within Patients
trum emitted by the tube. Although anatomic the uptaking tissue, where mT is the tissue mass We retrospectively considered 40 consecutive
information can be either assumed from gen- and mI is the iodine mass accumulated. The simu- patients who underwent both unenhanced and
eral models or obtained from the patient’s lative study was conducted for the thyroid, liver, contrast-enhanced thoracoabdominal CT in our
CT images, the actual quantity of iodine ac- spleen, pancreas, and kidneys, and the values of study. For each patient, the contrast-enhanced CT
cumulated by an organ during scanning re- these constants are shown in Table 1. examination was performed immediately after the
mains unknown, unless it is correlated to an Furthermore, we measured the relationship be- unenhanced CT, using the same scanning param-
in vivo measurable quantity, such as attenua- tween the iodine concentration and the correspond- eters and with the patient in the same position with
tion measured in Hounsfield units [24]. ing increment in Hounsfield units (ΔHU) resulting arms over the head.
In a recent phantom study, we described from the difference between the Hounsfield unit All patients were studied using a 16-MDCT
a method for evaluating the relationship be- value of the iodinated tissue and the Hounsfield scanner (Somatom Definition, Siemens Health-
tween organ radiation dose increase and the unit value of native tissue. For this purpose, we care), the same as that used in our previous study
difference in tissue attenuation as measured performed a series of measurements on known di- [25], operating at 120 kV in a helical mode with a
by Hounsfield units between two CT scans lutions of iodine contrast medium exposed to our reconstruction slice thickness of 7 mm and a slice
obtained before and after contrast medium 16-MDCT scanner (Somatom Definition, Siemens gap of 7 mm. Iodinated contrast medium (iopro-
administration [25]. Our method used a se- Healthcare), operating at 120 kV. We showed that, mide; Ultravist 300, Bayer Healthcare) containing
ries of measurements of iodine solutions to for a tissue with a density of ρT (g cm−3): 300 mg I/mL was administered IV at a dose of 1.5
correlate iodine concentration with Houns- mL/kg of patient weight, and contrast-enhanced
φI
field unit increment and a Monte Carlo simu- ∆HU = ρTγ (2). CT was performed during the portal phase (80
1 − φI
lation of an anthropomorphic phantom with seconds after the start of contrast administration).
iodinated organs to correlate iodine concen- The parameter γ(g−1 cm3) is defined as follows: We calculated the Hounsfield unit increment by
tration with the corresponding radiation dose µI
drawing the same circular region of interest (ROI)
increment. Thus, for the thyroid, liver, pan- γ=ρµ × 103 (3), in a well-defined region of each parenchyma for
I H 2O
creas, spleen, and kidneys, it was possible to both unenhanced and contrast-enhanced CT. For
determine the relationship between the mea- where γ is dependent on the x-ray energy spec- each patient, ROIs were drawn by the same radi-
sured Hounsfield unit increase and the corre- trum, so that it is constant for a given CT scanner ologists with 20 years of experience in CT. The
sponding radiation dose increment. operating at a definite kilovoltage value. Anyone ROIs were placed in homogeneously uptaking re-
The aim of the present work was to quan- gions of tissues, avoiding lesions, vessels, fat in-
tify the dose increment in CT due to the ad- TABLE 1: Organ Constants clusions, or other abnormalities.
ministration of iodinated contrast medium in The thyroid was sampled by drawing an ellipti-
Constant
the organs that accumulate the most contrast cal-shaped ROI with an area of 4.6 mm2 in either
medium by applying our simple experimental Organ a b the right or left lobe in a section passing through
method on a sample population undergoing Liver 38.64 0.8294 the middle part of the gland.
routine contrast-enhanced CT examinations. Spleen 80.81 0.9090 Two ROIs were drawn in the liver, one of 100
mm2 in the caudate lobe and the other 150 mm 2 in
Kidney 99.12 0.9162
Materials and Methods segment IV in an axial plane passing through the
Model Pancreas 136.35 0.9380 portal vein. According to recent literature [26], we
The method presented in our previous work Thyroid 111.07 0.9545 first considered fatty liver involution qualitative-
[25] allows the average dose increment in selected Note—Constants were derived from a study by ly, on the basis of a visual comparison of the at-
organs due to the administration of iodinated con- Amato et al. [25]. tenuation of hepatic parenchyma and vessels and
then quantitatively when hepatic attenuation val- In the present work, we applied this method
Increase (%) ues were lower than 48 HU and when the hepat- in a clinical setting and evaluated 40 consec-
Relative
Dose
71
utive patients who underwent CT before and
11
ic attenuation index (liver attenuation divided by
spleen attenuation) was less than 0.8. after contrast medium administration. The re-
The spleen was sampled near the hilum with sults are in agreement with our previous data,
Kidneys
Increment
an ROI of 100 mm 2. The kidney was sampled by confirming an increase in organ radiation dose
Downloaded from www.ajronline.org by 114.125.93.254 on 11/18/19 from IP address 114.125.93.254. Copyright ARRS. For personal use only; all rights reserved
139
HU
drawing an ROI of 100 mm 2 in the equatorial re- in contrast-enhanced CT compared with un-
24
TABLE 2: Mean Basal Hounsfield Unit (HU) Values and Increment and Relative Dose Increase for Five Organs Studied in 40 Patients
gion less subject to fibro-fatty involution, and the the kidneys showed the maximum among the
HU
34
3
head-tail passage with an ROI of 50 mm 2, which average dose increments (71%, resulting from
is the first region to undergo fibro-fatty involution. an attenuation increment of 139 HU). High re-
Increase (%)
Examples of the placement of ROIs are shown nal enhancement is, in fact, due to both their
Relative
Dose
Increment
Where two measurements were taken, as for lar, the level of contrast medium within renal
tubules can be up to 50–100 times higher than
HU
value was calculated and considered. Such values that in the blood because of the mechanisms of
were inserted into Equation 4 to estimate the per- tubular concentration and secretion [28].
Basal
centage dose increase in the five organs studied. The thyroid parenchyma showed the sec-
HU
36
10
Average values and related standard devi- crement of 41%, resulting from a Hounsfield
Relative
Dose
ations are reported in Table 2, whereas indi- unit increment of 87%. It should be noted,
33
8
vidual results for all 40 patients are shown in however, that these average values were char-
Table S2, which can be seen in the AJR elec- acterized by a broad dispersion, as shown in
Spleen
tronic supplement to this article (available Figure 1A and confirmed by the SD. This was
Increment
at www.ajronline.org). Five of forty patients due to the dispersion (SD of basal Hounsfield
HU
20
71
(patients 7, 20, 21, 30, and 32) presented with unit level, 13) of Hounsfield unit values in un-
hepatic steatosis, diagnosed on the basis of enhanced CT of the thyroid tissue as well as
Basal
the criteria described in the previous section. the different degree of vascularization and io-
HU
48
4
For each patient, the mean Hounsfield unit dine uptake among the patients.
value in unenhanced CT, the Hounsfield unit We noticed that the increase in the dose for
Increase (%)
increment after contrast administration, and the thyroid depended on tissue density on un-
Relative
Dose
the relative percentage increases in the radia- enhanced CT. As shown in Figure 2, denser
19
3
tion dose absorbed for the five organs studied thyroids showed a lower increase in attenuation
are reported in Table S2. and, consequently, in the radiation dose. More-
Liver
The average dose increments were 19% over, we know from the literature that paren-
Increment
for the liver, 71% for the kidneys, 33% for the chymal gland attenuation is proportional to io-
HU
49
10
spleen and pancreas, and 41% for the thyroid. dine concentration both in the normal thyroid
Note—See also Table S2 for individual values for all 40 patients.
Histograms of the relative increase in radiation and in diffuse thyroid disease [29]. There is
Basal
dose for each organ are presented in Figure 1. also a linear correlation between iodine uptake
HU
50
11
method for exploiting a series of measure- and by Doppler color-flow sonography [30, 31].
14
41
ments on phantoms of water solutions with This linear correlation can explain the inverse
varying iodine contents to correlate iodine correlation between native thyroid density and
Thyroid
concentrations with Hounsfield unit incre- dose increment because a thyroid with lower
Increment
ments. We also used a Monte Carlo simula- iodine concentration is more vascularized and
HU
87
31
tion of an anthropomorphic phantom with thus exhibits a higher Hounsfield unit incre-
iodinated organs to correlate iodine concen- ment during contrast-enhanced CT.
Basal
trations with the corresponding dose incre- Both the liver and spleen are richly vascu-
HU
75
13
ments, which revealed a close relationship larized; consequently, average Hounsfield unit
between the increase in Hounsfield units and increments of 49 and 71 HU, respectively, and
Value
the corresponding dose increments for the corresponding average dose increments of
Mean
SD
thyroid, liver, pancreas, spleen, and kidneys. 19% and 33%, respectively, were found.
Thyroid Liver
12 16
14
10
12
8
No. of Patients
No. of Patients
Downloaded from www.ajronline.org by 114.125.93.254 on 11/18/19 from IP address 114.125.93.254. Copyright ARRS. For personal use only; all rights reserved
10
6 8
6
4
4
2
2
0 0
0 20 40 60 80 100 0 5 10 15 20 25 30 35 40
Change in Radiation Dose (%) Change in Radiation Dose (%)
A B
Spleen Pancreas
14 12
12
10
10
8
No. of Patients
No. of Patients
8
6
6
4
4
2 2
0 0
0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60
Change in Radiation Dose (%) Change in Radiation Dose (%)
C D
Kidneys
16
14
12
No. of Patients
10
0
0 20 40 60 80 100 120
Change in Radiation Dose (%)
E
It is noteworthy that, in the liver parenchy- exhibit a dose increment different from the aver- (71 and 49 HU, respectively). This was due to
ma, unlike the thyroid gland, no relationship age value of the other patients without steatosis. the unequal speed of the increase in average
was found between the unenhanced attenuation The spleen and pancreas showed the same dose increment versus Hounsfield unit incre-
and radiation dose increment. In particular, the average dose increments (33%) but deriv- ment for these two parenchymatous tissues
five patients affected by liver steatosis did not ing from different Hounsfield unit increments caused by the anatomic differences in shape
x-ray contrast agents in computed tomography on dose increase in CT with iodinated contrast me- of thyroid iodine by CT. J Comput Assist Tomogr
the induction of dicentrics and gamma-H2AX dium. Med Phys 2010; 37:4249–4256 1991; 15:287–290
foci in lymphocytes of human blood samples. 26. Ma X, Holalkere NS, Kambadakone A, Mino-Ke- 30. Erbil Y, Ozluk Y, Giriş M, et al. Effect of Lugol
Phys Med Biol 2009; 54:6029–6039 nudson M, Hahn PF, Sahani DV. Imaging-based solution on thyroid gland blood flow and mi-
23. Goldman LW. Principles of CT: radiation dose quantification of hepatic fat: methods and clinical crovessel density in the patients with Graves’ dis-
and image quality. J Nucl Med Technol 2007; applications. RadioGraphics 2009; 29:1253–1277 ease. J Clin Endocrinol Metab 2007; 92:2182–
Downloaded from www.ajronline.org by 114.125.93.254 on 11/18/19 from IP address 114.125.93.254. Copyright ARRS. For personal use only; all rights reserved
35:213–225 27. Thomsen HS, Morkos SK. Contrast media and the 2189
24. Brooks RA. A quantitative theory of the Houns- kidney: European Society of Urogenital Radiology 31. Ansaldo GL, Pretolesi F, Varaldo E, et al. Doppler
field unit and its application to dual energy scan- (ESUR) guidelines. Br J Radiol 2003; 76:513–518 evaluation of intrathyroid arterial resistances dur-
ning. J Comput Assist Tomogr 1977; 1:487–493 28. Wong GT, Irwin MG. Contrast-induced nephrop- ing preoperative treatment with Lugol’s iodide
25. Amato E, Lizio D, Settineri N, Di Pasquale A, athy. Br J Anaesth 2007; 99:474–483 solution in patients with diffuse toxic goiter. J Am
Salamone I, Pandolfo I. A method to evaluate the 29. Imanishi Y, Ehara N, Mori J, et al. Measurement Coll Surg 2000; 191:607–612
F O R YO U R I N F O R M AT I O N
A data supplement for this article can be viewed in the online version of the article at: www.ajronline.org.