| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Interventional and Diagnostic Adaptative Imaging, INSERM ERI 13, Nancy
University, Centre Hospitalier Universitaire Nancy (TD4), Rue du Morvan, 54511
Vandœuvre-les-Nancy CEDEX, France.
2 Department of Radiology, Service d'Imagerie Guilloz, Centre Hospitalier
Universitaire, Nancy, France.
3 Siemens Medical Solutions, Saint Denis, France.
4 Cristal Laboratory, Necker Hospital, Paris, France.
5 Department of Urology, Centre Hospitalier Universitaire, Nancy, France.
Received August 25, 2007;
accepted after revision October 3, 2007.
Address correspondence to R. Grosjean
(r.grosjean{at}chu-nancy.fr).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. The stones were placed in a jelly simulating the X-ray attenuation of the kidneys. A dynamic platform was used to apply to the phantom free-breathing motion (sinusoidal motion in z-axis) and motion due to lack of maintenance of a breath-hold (5 mm·s–1 in z-axis). Determination of the chemical composition was performed with mean CT attenuation values obtained at 80 and 120 kV and with dual-energy CT attenuation values.
RESULTS. Two hundred forty-one human urinary stones were classified into six groups: uric acid, cystine, struvite, weddellite (calcium oxalate dihydrate), whewellite (calcium oxalate monohydrate), and brushite. With no motion, the use of dual energy enabled differentiation of all of the types of stones with statistically significant differences. Uric acid (–20 ± 22 H), cystine (106 ± 19 H), struvite (271 ± 16 H), weddellite (323 ± 5 H), brushite (415 ± 30 H), and whewellite (510 ± 17 H) were identified as distinct groups. Motion-induced mean CT attenuation values were significantly different from those obtained with no motion. With motion, dual-energy CT attenuation values did not allow differentiation of all stone types.
CONCLUSION. Dual-energy CT attenuation values can be used to predict the chemical composition of stones in vitro. However, when slight motion is applied to renal stones during image acquisition, the values become significantly different from those obtained with no motion. Consequently, confusion arises in differentiating stone types. A perfect breath-hold has to be performed for in vivo use of attenuation value to discern stone type.
Keywords: chemical composition • CT attenuation value • nephrolithiasis • renal stones • respiration motion
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Precise determination of the symptoms, localization, size, and chemical composition of stones is key to diagnosis and choice of therapy [11]. Extracorporeal shock wave lithotripsy is the most commonly used technique in the management of urinary stones, but the success of this treatment depends on the chemical composition of the stone and its corresponding fragility [12–15]. For example, brushite, cystine, and calcium oxalate monohydrate stones are more resistant to extracorporeal lithotripsy [16] than are the other types of stones we describe. Failure of extracorporeal shock wave lithotripsy increases medical costs, necessitates alternative treatment, and results in undesirable exposure of the renal parenchyma to shock waves. For these reasons, pretreatment determination of the composition of stones is essential. Since the early 1980s, studies have been conducted to determine stone composition on the basis of X-ray attenuation of stones in vitro [17–22] and in vivo [23–25].
Organs in the upper part of the abdomen, including the liver, kidneys, and spleen, move considerably as a result of respiration [26]. McCollough et al. [27], Alfidi et al. [28], and Ritchie et al. [29] have proved that physiologic motion decreases the quality of CT images. To our knowledge, no studies have been conducted to evaluate the influence of residual respiratory motion on the CT attenuation values of stones. The aim of our study was to determine with CT attenuation values the chemical composition of human renal stones in a jelly phantom and to analyze the influence of respiratory motion on stone classification.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Phantom and Dynamic Platform
The stones were placed in a jelly made of water, iodine, and animal
proteins (Fig. 1). The iodine
and protein concentrations were empirically chosen to ensure the jelly had an
X-ray attenuation similar to that of human kidney (30 H at 120 kV). To 1 L of
water, we added 21.6 g of animal protein and 0.01 mg of iodine. Each layer of
jelly, containing all the stones of one type, was successively settled in a
plastic container (280 x 210 x 110 mm). The jelly phantom was
homogeneous (30 ± 3 H). Stones were embedded in a layer 3 cm thick. The
jelly phantom had six layers, for a total thickness of 18 cm. The plastic
container was placed in a water tank. This water tank containing the six
layers of jelly and the 241 stones was placed on a dynamic platform designed
to accurately simulate respiratory motion in the z-axis
[30].
|
Applied Motion
Two motions were used: a simulation of free-breathing motion and a
simulation of motion due to lack of maintenance of a breath-hold. For free
breathing, a sinusoidal motion, described by Lujan et al.
[31], was applied with an
amplitude of 35 mm and a period of 5 seconds, as proposed by Moerland et al.
[32] and proved by Pasquier et
al. [33]. To achieve
independence from the phase of free-breathing motion, we performed three
acquisitions with free-breathing motion and averaged the data obtained. For
breath-hold motion due to lack of maintenance of a breath-hold, the phantom
was translated in the longitudinal axis (z-axis) with a speed of 5
mm/s. To simulate a perfect breath-hold, the platform was kept at rest.
CT Parameters
MDCT (Somatom Sensation 16, Siemens Medical Solutions) was performed at 80
and 120 kV, 200 mAs, 0.5-second gantry rotation time, 0.75-mm collimation, and
0.7-mm index of reconstruction. These parameters were those used in a typical
abdominal examination protocol.
Image Analysis
For measurement of CT attenuation values at image analysis, we devised an
interface based on Matlab (Mathworks). Stones were segmented from multiplanar
reformation by use of standard morphologic image-processing operations (global
threshold of 155 H, opening to remove pixel inferior in relation to three
pixels and closing to gather the contiguous zones separated by the
thresholding). For each acquisition and for each stone, the largest region of
interest (ROI) was set closest to the largest area of the stone
(Fig. 2). The size and the
position of the ROIs had been validated twice by an experienced radiologist
using a conventional soft-tissue window. A conventional soft-tissue window
(width, 350 H; level, 40 H) was used to record the mean ± CT
attenuation values within the ROI.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
At 80 kV, only uric acid stones were identified as a distinct group on the basis of CI. Struvite and cystine stones were not differentiated, but these two groups were differentiated from uric acid, calcium oxalate, and brushite stones. It was difficult to separate whewellite from weddellite and brushite stones (Fig. 3A). At 120 kV, cystine stones were clearly identified, but with this energy, uric acid stones were not differentiated from struvite stones. For whewellite, weddellite, and brushite stones, overlap was present but less important than at 80 kV (Fig. 3B).
CT attenuation values were lower at 120 kV than at 80 kV (Table 2). On average, the CT attenuation values at 120 kV were equal to 75% (± 17 kV) of the value at 80 kV. The differences between mean CT attenuation values were particularly high for whewellite stones (61%) and struvite stones (63%). Despite a larger spread of CT attenuation values at 80 kV, there was less overlap at 120 kV (Fig. 3A, 3B). Only cystine stones at 120 kV and uric acid stones at 80 kV were identified with statistical significance. However, the dual-energy CT attenuation values obtained at 80 and at 120 kV allowed statistically significant differentiation of all types of stones (Fig. 4). Uric acid (–20 ± 22 H), cystine (106 ± 19 H), struvite (271 ± 16 H), weddellite (323 ± 5 H), brushite (415 ± 30 H), and whewellite (510 ± 17 H) were identified as distinct groups.
|
|
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In 1998, Mostafavi et al. [20] asserted, having studied only 102 stones, that the best CT parameter for accurately determining the chemical composition of stones was the mean CT attenuation value obtained at 120 kV. This affirmation was in accord with our results. At 120 kV the overlap between types of stones was less, and it was easier to differentiate the stone groups. Despite the overlap problem, it was possible to classify the types of stones according to mean CT attenuation value and 95% CI. At 120 kV, from least to most dense, the stones types were as follows: uric acid (437 ± 45 H), struvite (461 ± 117 H), cystine (625 ± 23 H), whewellite (797 ± 94 H), weddellite (1,017 ± 201 H), and brushite (1,216 ± 150 H). In most of the previous studies, the chemical composition of the stones had been assessed only at 120 kV. This kilovoltage is the one most commonly used for clinical abdominal imaging. Consequently, only our results obtained at 120 kV can be compared with those of others. Our classification is similar to those of Mitcheson et al. [17], Bachmann et al. [34], and Mostafavi et al. [20]. Hillman et al. [18], New-house et al. [19], Saw et al. [21], Bellin et al. [22], and Motley et al. [23] switched cystine and struvite (Table 3). This difference can be explained by the fact that there is often a large overlap between the CI of struvite and the CI of cystine in these studies. Deveci et al. [35] switched whewellite and weddellite, but they used only one weddellite stone, opening the finding to statistical criticism.
|
CT attenuation values were 75% higher at 80 kV than at 120 kV, and the classification obtained at 80 kV was not the same as that at 120 kV, cystine and struvite being switched. This variation in CT attenuation value determination according to kilovoltage has been described by McKetty [36] and Bushberg [37]. Without motion, for all stones, findings on single-energy scanning (80 or 120 kV) did not allow accurate differentiation of the chemical composition of stones. Dual-energy scanning, however, provided additional information on chemical characterization. As shown by Mostafavi et al. [20], dual-energy CT attenuation values are extremely reliable in differentiating all stones. Even if Mostafavi et al. had been the only investigators to study the six most frequent types of stones (Table 3), they conducted their study without considering renal X-ray attenuation, and the number of stones used could not guarantee the statistical significance of the results.
In our study, the use of the dual-energy CT attenuation values made it possible to characterize all renal stones with statistical significance. In clinical use, imaging for nephrolithiasis is generally performed at 120 kV. Because the received dose is proportional to the square of the kilovoltage, use of this technique can lead to an approximately 40% increase in received dose, possibly less. The second acquisition thus can be minimal (centered on the stone without acquisition of images of the rest of the abdomen) and performed with low energy (80 kV) [38–40]. This additional dose can be a problem for children but can be justified for adult patients because accurate and fast determination of the chemical characterization of renal stones can lead to quicker and more accurate treatment. This new protocol can be facilitated with the dual-source CT (Somatom Definition, Siemens AG Medical Solutions).
Several studies have been conducted in the attempt to determine the chemical composition of renal stones on the basis of in vitro and in vivo X-ray attenuation of the stones (Table 3). The in vitro studies, however, did not reproduce normal abdominal wall and fat, perinephric fat, or the spine, causing uncertainty about standardization of the values obtained. For example, Bellin et al. [22] used excised pig kidney placed in water, Mostafavi et al. [20] and Deveci et al. [35] placed stones in air, and Saw et al. [21] placed stones in water. The influence of the surrounding media on CT attenuation values has been proved by Grosjean et al. [41], who found the mean CT attenuation values of stones vary with the surrounding media tested (air, water, and jelly). This finding can be explained by the beam-hardening effect and insufficient correction of the effect with CT algorithms [42]. In our study, we used a jelly made of water, iodine, and animal proteins and surrounded it with water. The CT attenuation value of the jelly (30 ± 6 H at 120 kV) allowed us to reproduce renal attenuation as accurately as possible, but the phantom did not reproduce exactly normal abdominal wall and fat.
The influence of the surrounding media on mean CT attenuation value explains the difference between our values and those obtained in the other studies [20, 21, 35], which did not respect the normal absorption of the abdomen because the stones were scanned within an air-filled environment or in water. For example, to prevent overlapping densities and to avoid the absorption of X-ray beams by water or fat, Deveci et al. [35] used an air-filled environment instead of phantoms containing water or fat. Thus the density differences increased, and overlap did not occur. Consequently, because the in vitro conditions were too far from the in vivo conditions, the results cannot be considered a reference for in vivo determination of chemical composition.
Because our study had, to our knowledge, the largest number of stones described in the literature to date, the results have statistical significance for the six most frequent types of stones. Only stones containing at least 85% of one component and having low content of a secondary component were used. These stones can legitimately be considered to reflect the behavior of the principal component and in consequence can be considered pure. Overrepresentation of some components (for example, calcium oxalate monohydrate and brushite) was deliberate because of their particular resistance to extracorporeal shock wave lithotripsy. Carbapatite stones were not included in our study because they are rarely pure and almost always are multiphasic with high content of a secondary component. It was also difficult to collect enough pure carbapatite stones to obtain significant results. Pure carbapatite stones are, in practice, less frequent than cystine stones. Our stones were selected from the purest possible stones. Even though most stones are polycrystalline, it was necessary to work with pure stones or stones considered pure. Because differentiation of pure stones has not been proved, study of mixed stones would not be realistic.
Our phantom was closer to in vivo conditions than phantoms used in in vitro studies of the X-ray attenuation of the kidney. With this phantom and the large number of stones, we attempted to prove that dual-energy CT attenuation values can be clinically useful for determining the chemical composition of pure renal stones. However, even if a breath-hold can generally be maintained over the acquisition time, for some patients (e.g., those with sharp pain due to renal colic, children, and dyspneic patients), a breath-hold is particularly difficult [43]. To analyze the influence of motion on the X-ray attenuation of stones, two types of motion were applied to the phantom: free-breathing motion and motion due to lack of maintenance of a breath-hold. We found that CT attenuation values depend on motion. With the two types of movement, CT attenuation values generally decreased. With either type of motion, the mean CT attenuation values were significantly different from those obtained at rest. Therefore, even dual-energy CT attenuation values cannot be used to determine the chemical composition of renal stones in the presence of motion.
The explanation for the influence of motion is the partial volume effect due to motion. With 16 x 0.75 mm collimation and a gantry rotation time of 0.5 second, 24 mm are scanned per second, but in that time, the stones have moved at least 5 mm with lack of maintenance of a breath-hold. The partial volume effect occurs during imaging of any part of the body in which the anatomic relations are changing rapidly in the z direction [44]. To keep the partial volume effect at a minimum, the thinnest available slice thickness has to be used. In addition, acquisition time generally increases when slice thickness is decreased. The longer the acquisition time, the greater is the partial volume effect due to the presence of motion. Because of this partial volume effect paradox, a compromise between slice thickness and acquisition time is needed. To minimize the acquisition time, gantry rotation time can be decreased. It is noteworthy that the gantry rotation time of 500 milliseconds did not allow minimization of motion artifacts but corresponds to the abdominal imaging protocol in current clinical use. Even with the fastest available gantry rotation time (420 milliseconds), it is impossible to remove motion artifacts. Ritchie et al. [29] found that the gantry rotation time necessary to eliminate motion artifacts caused by quiet breathing was 93.5 milliseconds.
Our results indicate that determination of the chemical composition of renal stones has to be done with images obtained during a perfect breath-hold. We can never be sure, however, that a patient has realized a perfect breath-hold. Imprecision due to motion is a limitation of CT determination of the chemical composition of renal stones. Without motion, prediction of renal stone composition with dual-energy CT attenuation values is reliable. When motion is applied, the technique becomes unusable.
Saw et al. [21] scanned 127 urinary stones of known composition placed in a water bath (120 kV, 240 mA). By referring to the model of Hu and Fox [45], which showed that measurement of the CT attenuation value of an object with helical CT is affected by collimation width and pitch, Saw et al. found that the measured attenuation of stones declined with increasing collimation width owing to the partial volume effect. In addition, scanning 63 stones in a waterbath (120 kV, 100 mA), Hill-man et al. [18] found that even if it is clear that small stones will be subject to partial volume inaccuracies in measurement of CT attenuation and that this source of inaccuracy can be further accentuated by the respiratory movement of a patient, CT may prove a valuable adjunct to traditional laboratory and clinical methods of establishing the chemical composition of stones. To avoid partial volume inaccuracies, we chose the thinnest possible slice thickness for our study (0.75 mm). It is essential to use a narrow slice width to ensure the accuracy of attenuation values in helical CT [46]. But decreasing the slice thickness has two opposing effects. Stone detection can be improved with a decrease in partial volume effect, but it can hampered by an increase in noise [47–49]. In addition, choosing table feed values greater than the nominal section thickness increases degradation of the slice sensitivity profile [50]. For lesions smaller than the section thickness, a reduction in contrast enhancement due to the partial volume effect can be observed.
There may be variability in CT attenuation values of scans obtained with different CT scanners manufactured by different companies and even among different scanners made by the same manufacturer and of the same model [51, 52]. This phenomenon may explain the slight variations in CT reports of attenuation values. In addition, most renal stones in humans are not pure. Heterogeneity can be very important, and the proportion of constituents can vary considerably [46, 53, 54]. To determine absolute CT attenuation values, Mostafavi et al. [20] used 1-pixel ROIs to measure CT attenuation. Deveci et al. [35] used three 0.01-cm2 ROIs. In applying these methods, the investigators did not take into account the structural heterogeneity of mixed calculi. Because of this structural heterogeneity, we used the largest ROI within a stone to obtain a reliable mean CT attenuation value for each stone, as did Motley et al. [23]. When motion is applied, however, the larger the ROI, the more pixels from the environment of the stone can be included and lead to an increase in partial volume effect.
A limitation of our study was that only stones with a diameter between 7 and 25 mm were assessed. The influence of stone size on CT attenuation values was not studied. The results can be degraded when the diameter of the stones approaches the slice thickness. The effect of stone size will be the subject of future work.
In conclusion, with single-energy CT, overlap between types of renal stones makes it difficult to reliably determine the chemical composition. Dual-energy CT attenuation values can be used for accurate prediction of stone composition in vitro. When slight motion is applied to renal stones during acquisition, however, CT attenuation values and even dual-energy CT attenuation values became significantly different from those obtained at rest and consequently can lead to confusion between stone types. Therefore, for in vivo application of this technique, a perfect breath-hold has to be performed by the patient, even during MDCT.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  L.-J. Zhang, Y.-E Zhao, S.-Y. Wu, B. M. Yeh, C.-S. Zhou, X.-B. Hu, Q.-J. Hu, and G.-M. Lu Pulmonary Embolism Detection with Dual-Energy CT: Experimental Study of Dual-Source CT in Rabbits Radiology, July 1, 2009; 252(1): 61 - 70. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. T. Boll, N. A. Patil, E. K. Paulson, E. M. Merkle, W. N. Simmons, S. A. Pierre, and G. M. Preminger Renal Stone Assessment with Dual-Energy Multidetector CT and Advanced Postprocessing Techniques: Improved Characterization of Renal Stone Composition--Pilot Study Radiology, March 1, 2009; 250(3): 813 - 820. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
