| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 All authors: Department of Radiological Sciences, David Geffen School of Medicine, University of California, Los Angeles, Medical Plaza 200, Ste. 165-59, Los Angeles, CA 90095.
Received January 6, 2008;
accepted after revision May 9, 2008.
C. Pankin is an employee of Siemens Medical Solutions.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Eleven healthy volunteers and seven clinically referred patients (10 men, eight women; mean age, 43.1 years; range, 20–71 years) underwent conventional single-energy CT and dual-energy CT examinations of both lower extremities with a dual-source CT scanner. Dual-energy reconstructions were made at combined tube voltages approximating 86, 98, 110, 122, and 134 kVp. Quantitative and qualitative analyses were performed on six tendons in each lower extremity, and the findings were compared with single-energy CT findings. The radiation dose involved was recorded in each case.
RESULTS. A trend toward increasing tendon attenuation was observed with increasing reconstructed tube voltage. The group of single-energy CT reconstructions proved significantly superior to each of the dual-energy CT reconstructions with regard to signal-to-noise ratio (F = 35.25, p < 0.0001) and contrast-to-noise ratio (F = 37.19, p < 0.0001), although interobserver agreement in subjective ranking was poor. Dual-energy CT had a significantly higher radiation dose (p < 0.05) than single-energy CT.
CONCLUSION. Dual-energy CT of lower-extremity tendons, irrespective of the reconstruction tube voltage chosen, yields multiplanar reformations inferior to those of single-energy CT with regard to signal-to-noise and contrast-to-noise ratios while involving significantly escalated patient exposure to ionizing radiation. Whether the tissue-differentiating promise of dual-energy CT is realized in future studies and warrants such concessions remains to be seen.
Keywords: dual-energy CT • ionizing radiation • tendon
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Little evidence of the feasibility of dual-energy CT in the evaluation of tendinous continuity and integrity exists in the literature. A report by Johnson et al. [5] suggested that dual-energy CT is not associated with marked increases in ionizing radiation exposure of patients. But what might dual-energy CT have to offer in imaging of lower-extremity tendons? Because the attenuation responses of various tissues differ, dual-energy CT should, at least in theory, enable isolated depiction of tendons without interference from adjacent soft tissues. In the presence of trauma, which results in peritendinous edema, dual-energy CT, owing to confident separation of tendon from adjacent fat stranding, may be useful for evaluation of the integrity of tendon. Similar evaluation of isolated tendon may be possible in the presence of degenerative tenosynovitis, in which separation of tendon from adjacent sheath fluid may not be possible with conventional CT. The utility of dual-energy CT in these roles remains to be proved.
Even if dual-energy reconstructions prove useful in clinical practice, diagnostic interpretation of dual-energy CT scans will rest with the cross-sectional axial, sagittal, and coronal multiplanar reconstructions (MPRs) produced, as is the case with other 3D CT applications. Dual-energy CT is unique in that once the data have been obtained, retrospective reconstruction can be achieved at any desired combined tube voltage between 80 and 140 kVp, allowing optimization of tube voltage to suit the tissue of greatest interest without repeated patient scanning.
We hypothesize that dual-energy CT is inferior to single-energy CT with regard to diagnostic quality of MPRs in tendon evaluation and that it results in higher radiation exposure. The basis for this hypothesis is that according to the fundamentals of radiation physics, image contrast is reduced with increasing tube voltage. As a result, one can expect the contrast-to-noise ratio (CNR) to be lower for dual-energy CT examinations, which involve at least a partial contribution from a CT tube operated at 140 kVp, compared with single-energy CT, which is performed at 120 kVp. Similarly for signal-to-noise ratio (SNR), a greater proportion of photon–tissue interactions occur in the form of coherent scatter rather than the photoelectric effect with increasing tube voltage. The result is a decrease in SNR at 140 kVp compared with 120 kVp, although one would expect this effect to be offset at least in part by relatively low noise from the 80-kVp tube–detector pair.
To establish an optimal voltage at which MPRs should be evaluated for diagnostic interpretation, should our hypothesis be proven incorrect, we sought to evaluate whether any one particular combined reconstruction voltage is associated with superior tendon depiction on dual-energy CT MPRs. Should our hypothesis hold true, voltage determination will represent a considerable obstacle to the future study and implementation of dual-energy CT, such use depending on the tissue-differentiating capabilities of the technique.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Initial processing was performed on the CT console with incorporated software. A fellowship-trained musculoskeletal radiologist with 7 years of experience was responsible for all data postprocessing. The raw data acquired at single-energy CT were reconstructed at 20-cm2 field of view and 2-mm slice thickness at a 2-mm increments with a medium-smooth soft-tissue kernel (B30f). Processing of the dual-energy data involved integration of the data from each tube–detector pair, facilitating image reconstruction at any desired relative contribution of the low- and high-voltage tubes. Reconstructions were made at 20% intervals from 86 kVp (low voltage–to–high voltage ratio, 90%/10%) to 134 kVp (low voltage–to–high voltage ratio, 10%/90%), each 20% interval corresponding to an increase in mean voltage of approximately 12 kVp. Thus dual-energy reconstructions were performed at combined tube voltages approximating 86, 98, 110, 122, and 134 kVp. These data were reconstructed with a 20-cm2 field of view and 2-mm slice thickness at 2-mm increments with a soft-tissue kernel (D30f) similar to that used for single-energy reformatting. No single kernel on the CT system was available for both single- and dual-energy CT reconstruction, so the most-similar soft-tissue kernels were used.
Quantitative Analysis
Evaluation of tendon attenuation focused on six tendons within various
compartments of the ankle joint—the Achilles, tibialis posterior, flexor
digitorum longus, flexor hallucis longus, peroneus longus, and tibialis
anterior tendons. A region of interest was manually placed on each of these
tendons at two locations, 5 cm and 1 cm proximal to the ankle joint, and the
attenuation at each location was recorded. The region of interest measured was
as large as possible but sufficiently small to avoid inclusion of adjacent fat
and bone. Areas of beam hardening, found along the middle thirds of these
structures because of their proximity to dense cortical bone, were
intentionally avoided. The SD, which serves as a quantitative marker of the
image noise for the surrounding air, and attenuation of the gastrocnemius
muscle were measured in each case. The SNR and CNR for each tendon were
calculated according to the following formulas: SNR = mean tendon attenuation
/ SD noise(air) CNR = (mean tendon attenuation –
gastrocnemius attenuation) / SD noise(air).
In all, these measurements were derived from six tendons on each side in 18 subjects at single-energy and at five separate dual-energy tube voltages (86, 98, 110, 122, and 134 kVp), for a total of 1,296 SNR and CNR values, which formed the basis of subsequent quantitative statistical analysis.
Qualitative Analysis
Two board-certified musculoskeletal radiologists, not including the
radiologist responsible for data postprocessing, blinded to the method used
for data acquisition, including the voltages of dual-energy examinations,
performed an independent qualitative evaluation of the axial MPRs obtained in
each case. Data sets were randomly presented in a 6 x 1 screen format,
each with a 10 x 10 cm field-of-view, on a workstation (Wizard 3D,
Siemens Medical Solutions). Each of the reconstructed data sets was
subjectively ranked with regard to overall quality of tendon depiction from 1
(best) to 6 (worst), according to overall tendon conspicuity and edge
definition. Allocation of joint ranks was permitted in the event of comparable
image quality. Observers were requested to note artifacts identified in any
data set, irrespective of cause.
Radiation Exposure
The effective dose to each subject was recorded from the scanner console
for both the single- and dual-energy CT examinations in the form of CT dose
index (CTDI) [6]. The volume
CTDI, a derivative of the CTDI that can be used to express the average dose
delivered to a scan volume for a specific examination, was recorded, as was
the dose–length product.
Statistical Analysis
Statistical analysis was performed with Stata software (version 10.0,
Stata). Repeated measures two-factor analysis of variance was used to evaluate
for differences in SNR and CNR data obtained from the single-energy and each
of the dual-energy reconstructions. A within-subject design was used to
minimize the effect of intersubject variability. Friedman's nonparametric test
was used to verify that any significant results obtained did not occur as a
result of nonnormal distribution of data. The presence of significant
interobserver difference between the ranks assigned to each reconstructed data
set was evaluated with weighted kappa analysis, whereby a score of 1 was
assigned for corresponding tendon scores and 0.5 to scores that differed by a
single rank. Interobserver deviation of more than one rank was not assigned
any score according to the weighting system used. A kappa value of 0 was taken
as indicative of poor interobserver agreement; 0.01–0.20, slight
agreement; 0.21–0.40, fair agreement; 0.41–0.60, moderate
agreement; 0.61–0.80, good agreement; and 0.81–1.00, excellent
agreement. Intrarater agreement was evaluated with the Cronbach alpha
reliability coefficient. A paired Student's t test was used to
evaluate for potential difference in the dose–length product measure of
radiation exposure between single- and dual-energy examinations. A value of
p < 0.05 was designated as representing statistical significance,
prompting rejection of the null hypothesis.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
|
Qualitative Analysis
Observer 1 designated the single-energy CT MPRs of superior image quality
with a median ranked score of 1.0. For observer 2, the 134-kVp reconstructions
were most consistently awarded the top ranking, for a median ranked score of
1.5. Weighted kappa analysis revealed a poor level of interobserver agreement
between the two observers with regard to ranking of the various image data
sets (
= 0.014). Intrarater agreement was considerable for observer 1
(
= 0.85) and less so for observer 2 ( = 0.56, Cronbach alpha
reliability coefficient).
Radiation Exposure
Calculated mean dose profiles for single- and dual-energy CT are shown in
Table 4. A statistically
significant difference between the techniques was observed with regard to both
volume CTDI (p = 0.004) and dose–length product (p =
0.003).
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
Although the clear objective superiority of single-energy CT was not
indisputably found in the subjective analysis, one of the two observers did
score preferentially in favor of this technique. Nonetheless, the poor
interobserver agreement in ranks assigned ( = 0.014) and the relative
consistency with which the observers individually assigned their ranks
( = 0.85 for observer 1, = 0.56 for observer 2) suggest that
any subjective difference that might exist between single- and dual-energy CT
data may be subtle. With regard to ionizing radiation, dual-energy CT was
determined to involve significantly higher levels of exposure per helical
rotation (volume CTDI, p = 0.004) than single-energy CT and as a
function of the entire examination (dose–length product, p =
0.003).
Despite considerable relatively recent enthusiasm about potential clinical applications, implementation of dual-energy CT is entering its fourth decade [7, 8]. Early attempts at application of this technique were limited by a number of technical obstacles, including insufficient tube current, the resultant variations in attenuation values, prolonged imaging times, and the requirement for sequential rather than simultaneous scans for complete dual-energy data acquisition [9–11]. The more recent past has seen remarkable developments in CT hardware and software, such that the versatility and implementation of this technique are now almost unrecognizable compared with those of its infancy [12]. It was, however, not until the introduction of dual-source CT in 2005 that application of these advances in dual-energy scanning rekindled interest in the technique. Because it allows simultaneous dual-energy MDCT data acquisition, dual-source CT addresses many of the limitations experienced in earlier attempts at application of the technique, facilitating the transition of dual-energy CT from theory to reality [1].
Early reports on the potential efficacy of the dual-energy approach focused on differentiation of iodine-based contrast medium–containing structures, such as vessels and enhanced viscera. This focus reflects the relatively exaggerated variation in attenuation values of high-atomic-number atoms, such as iodine, during imaging at varying tube voltages [5, 13]. The future of dual-energy CT is much less certain with regard to evaluation of collagenous tissues, such as tendon, in which relatively muted attenuation responses occur. In this study, during the transition from a dual-energy CT reconstruction voltage of 86 kVp to 134 kVp, the overall mean individual tendon attenuation variation was an increase of 2.49 HU (range, –14.2 to 20.6 HU), representing a mean increase in attenuation of 2.88% from baseline. This finding suggests that in the absence of marked alteration in composition, and thus attenuation response, of a particular tendon affected by a pathologic process, the role of dual-energy CT in the evaluation of lower-extremity tendons is far from certain.
There is little doubt that studies of the material-specific 3D applications of dual-energy CT will be performed in the near future, and the findings are eagerly awaited. Potential applications include isolated tendon depiction in the presence of trauma. In this use, differentiation of tendon from soft-tissue attenuation related to peritendinous edema or tenosynovitis may be possible, allowing confirmation of tendon integrity and identification of sites of tendinosis not otherwise appreciable with the limited contrast resolution of conventional single-energy CT. To our knowledge, there is no evidence to suggest that this technique is superior to single-energy CT for any of these applications. Our findings suggest that future studies should proceed with caution because use of the described dual-energy CT protocol necessitates concessions in image quality in the form of objective SNR and CNR and in patient exposure to ionizing radiation. Nonetheless, should such studies show a distinct diagnostic advantage of dual-energy CT over conventional single-energy CT in tendon analysis, particularly in patients for whom MRI is contraindicated (e.g., those with pacemakers or claustrophobia), the risk-to-benefit ratio may swing in the direction of the newer technique. In the absence of such contraindications, it is unlikely that dual-energy CT will challenge MRI, at least in its current form, as the reference standard technique for tendon evaluation.
We conclude that dual-energy CT of lower-extremity tendons, irrespective of the reconstruction voltage chosen, yields MPRs inferior to those obtained with single-energy CT with regard to SNR and CNR while entailing significantly escalated patient exposure to ionizing radiation. Whether the tissue-differentiating promise of dual-energy CT is realized in future studies and warrants the concessions remains to be seen. Such investigation should proceed with caution, however, in light of the findings of our study.
Acknowledgments
We thank Phil Ender, Xiao Chen, and Rose Medeiros, Statistical Consulting
Group, UCLA Academic Technology Services, for their considerable input.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
