| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Thoracic Imaging, Hospital Calmette, University Center of Lille,
Blvd. Jules Leclerc, 59037 Lille, France.
2 Department of Nuclear Medicine, Hospital Roger Salengro, 59037 Lille,
France.
3 Department of Pulmonology, Hospital Calmette, University Center of Lille,
59037 Lille, France.
4 Department of Medical Statistics, Hospital Calmette, University Center of
Lille, 59037 Lille, France.
Received July 11, 2005;
accepted after revision October 31, 2005.
Address correspondence to: M. Remy-Jardin
(mremy-jardin{at}chru-lille.fr).
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Forty-nine consecutively examined patients (30 men, 19 women; mean age, 59 years) with known or suspected right ventricular dysfunction secondary to bronchopulmonary (n = 30) or pulmonary vascular (n = 19) disease underwent ECG-gated 16-MDCT angiography of the heart (rotation time, 0.42 second; 120 kV; 300 mAs; collimation, 12 x 0.75 mm; pitch, 0.2) after CT angiographic examination of the entire thorax according to a standard protocol. Biphasic administration of a 30% contrast agent was systematically performed (phase 1, 90 mL at 3 mL/s; phase 2, 30 mL at 1.5 mL/s); no patient received additional medication. Right ventricular ejection fraction was calculated after two reviewers in consensus determined the reconstruction windows and segmentation of the right ventricular cavity on a series of diastolic and systolic short-axis images. The results were compared with those of equilibrium radionuclide ventriculography.
RESULTS. At data acquisition, the mean (± SD) heart rate of the study group was 82 ± 13.87 beats per minute (BPM) (range, 51-115 BPM). ECG showed a sinus rhythm in 30 (61%) of the patients and irregular cardiac rhythm in 19 (39%) of the patients. Agreement between the two techniques was estimated by intraclass correlation coefficient (0.77), the method of Bland and Altman (limits of concordance, -14.9 and 13.7), and percentage of variability between two measurements expressed by mean absolute percentage error (12.1%). The estimated effective dose for heart examination was 7.48 mSv with CT and 5 mSv with scintigraphy. The mean effective dose for the chest and heart CT examinations was 11.64 mSv.
CONCLUSION. Right ventricular ejection fraction can be reliably estimated with 16-MDCT in unselected patients.
Keywords: cardiopulmonary imaging • CT technique • heart • lung • MDCT
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
MDCT can be used for assessment of RV performance with the addition of retrospective ECG gating and use of dedicated cardiac analysis software. Most attention has been directed at validation of such tools for left ventricular functional analysis [20-23]. The purposes of this study were to assess RV ejection fraction (RVEF) with ECG-gated MDCT and to compare the results with those of equilibrium radionuclide ventriculography in a population of patients with respiratory disorders referred for CT angiography of the chest.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Before the introduction of MDCT technology, each patient referred for the management of chronic lung disease at our institution underwent unenhanced CT of the thorax. Because MDCT can give additional information about right heart function, inclusion of CT angiography was approved by our institutional review board and ethics committee in the clinical context of known or suspected RV dysfunction. Informed consent from patients was not required. No ß-blockers for reducing heart rate were administered before the examination. The mean heart rate of the study population was 82 ± 13.87 beats per minute (BPM) (range, 51-115 BPM). The cardiac rhythm was regular in 30 (61%) of the patients and irregular (i.e., presence of a few extrasystolic beats) in 19 (39%) of the patients.
CT Acquisition Parameters
CT examinations were performed with a 16-MDCT scanner (Sensation 16,
Siemens Medical Solutions). The CT protocol consisted of two successive
acquisitions, both in the craniocaudal direction, with the patients scanned in
the supine position and after deep inspiration. The first acquisition
consisted of non-ECG-gated CT angiography of the entire thorax and was called
the diagnostic scan (scanning parameters: 80-120 kV; 60-100 mAs; rotation
time, 0.5 second; collimation, 16 x 0.75 mm; pitch, 1.5). This
examination was systematically performed with online tube current modulation
(Care Dose, Siemens Medical Solutions). The second acquisition was focused on
the cardiac cavities and consisted of ECG-gated CT angiographic examination of
the heart (rotation time, 0.42 second; 120 kV; 300 mAs; collimation, 12
x 0.75 mm; pitch, 0.2) and was called the functional scan. Because
ECG-controlled dose modulation reduces the milliamperage during the systolic
phase, it could theoretically be responsible for degradation of systolic
images and thus influence the segmentation process at the level of the right
ventricle during systole. Consequently, no dose reduction system monitored
with the ECG tracing was used during this acquisition. The interval between
the two acquisitions varied between 8 and 10 seconds, including a minimum of 4
seconds between the two helical scans and the time needed to reposition the
table for the second acquisition. The patient was allowed to breathe between
the two scans. The dose-length product was systematically recorded after each
acquisition.
CT Injection Protocol
Biphasic administration of 120 mL of 300 mg I/mL contrast material
([iohexol], Visipaque 300, Amersham Health) was systematically performed. The
first phase was aimed at obtaining optimal vascular opacification for the
diagnostic scan and consisted of administration of 90 mL at 3 mL/s. The
automatic bolus-triggering software program available on the CT unit (Care
Bolus, Siemens Medical Solutions) was systematically used. A circular region
of interest was positioned at the level of the ascending aorta, and the
threshold for triggering data acquisition was preset at 100 H (standard chest
CT angiographic protocol in our department). The second phase was aimed at
obtaining enhancement of the cardiac cavities compatible with delineation of
the ventricular cavities during the functional scan (i.e., without streak
artifacts caused by too high a concentration of contrast material within the
right cardiac cavities). Images were obtained with administration of 30 mL of
contrast material at 1.5 mL/s immediately after administration of the first 90
mL of contrast medium for the diagnostic scan.
CT Image Reconstruction
The diagnostic scans were reconstructed according to the protocol used in
clinical practice. Contiguous 1-mm-thick images were viewed in both
mediastinal (window width, 450 H; window center, 80 H; soft reconstruction
kernel) and lung parenchymal (window width, 1,600 H; window center, -600 H;
high spatial frequency algorithm) window settings.
The functional scans were reconstructed as follows. The first step consisted of choosing the temporal window for reconstruction of systolic and diastolic images. To identify the maximal systolic contraction and diastolic relaxation phases, we acquired transverse test images in 5% steps through the entire R-R interval at the midventricular level. End-diastolic and end-systolic phases were identified from the ECG tracing and were controlled visually as the images showing the largest and smallest endocavitary diameters, respectively. The second step consisted of reconstruction of the volume scanned during the functional scan in the diastolic and systolic phases. Two sets of 1-mm-thick axial transverse scans obtained at 0.7-mm intervals were reconstructed with a soft kernel (B30f) on the basis of the temporal window previously chosen. The third step consisted of reconstruction of 3-mm-thick contiguous 2D images of the cardiac cavities along the short axis. Two series of images were then reconstructed: the diastolic and systolic short-axis images, which enabled subsequent calculation of functional parameters (Figs. 1A, 1B, 1C, and 1D). The first two steps of the retrospective ECG-gated image reconstruction were performed on the scanner workstation (Navigator, Siemens Medical Solutions). The third step of image reconstruction was performed on a commercially available console (Leonardo workstation, Siemens Medical Solutions). To improve the temporal resolution and to avoid motion artifacts, we used a multisegmental image reconstruction algorithm (with raw data from up to two cardiac cycles) (Adaptive Cardio-Volume algorithm, release VA 70, Siemens Medical Solutions) for generating images in each case.
|
|
|
|
|
|
Segmentation of RV cavity—The segmentation process required delineation of the RV endocardial contours and identification of the RV borders with the pulmonary trunk and right atrium. The endomyocardial contours were manually delineated on short-axis images of the right ventricle at end-diastole and end-systole (Figs. 3A, 3B, 3C, 3D, 3E, 3F, 3G, and 3H). Window settings were adapted for each examination to achieve the best contrast between myocardium and blood. Papillary muscles, the moderator band, and trabeculations of the right ventricle were included in the lumen. The RV outflow tract was included in the RV volume. Endocardial drawings were rated as precise when RV endocardial contours were easily delineated and as imprecise in cases of difficult delineation. In the latter situation, factors affecting the manual tracing of endocardial contours were systematically noted.
|
|
|
|
|
|
|
|
The junction between the right ventricle and the pulmonary trunk was recognized by means of visualization of the pulmonic valves or identification of the junction between the ventricular myocardium and the pulmonary trunk. The junction between the right atrium and the right ventricle was based on recognition of the junction between the interatrial and interventricular septa, identification of the atrioventricular valve, and/or identification of the right coronary artery in the atrioventricular septum.
Calculation of RVEF—RV function was evaluated on the Leonardo console with the manufacturer's software dedicated to analysis of cardiac function (Syngo Argus software, 2003 version; Siemens Medical Solutions). The analysis software calculated RV end-diastolic and end-systolic volumes by summing the volumes of all short-axis slices (Simpson method), RV stroke volume (RV end-diastolic volume minus RV end-systolic volume), and RVEF (RV stroke volume / by RV end-diastolic volume x 100%). RVEF results were compared with those of equilibrium radionuclide ventriculography performed within 24 hours.
The CT protocol was supervised by two reviewers, each of whom had 1 year of experience in cardiac CT at the time of initiation of this study. The reviewers determined in consensus the two operator-dependent steps, namely the choice of the reconstruction window and delineation of the endocardial contours of the right ventricle.
Equilibrium Radionuclide Ventriculography
All patients underwent equilibrium radionuclide ventriculography at rest in
the supine position. The procedure was performed with red blood cells labeled
in vivo with 20 mCi of 99mTc. Data were acquired in 45-degree left
anterior and in 30-degree right anterior oblique views with a gamma camera
(DST, SMV International). All studies were formatted at 16 frames per cardiac
cycle. R-R interval and heart rate were recorded. Cardiac cycles with R-R
intervals not within 20% of the average value were discarded. Left ventricular
ejection fraction and RVEF were determined with the equilibrium technique by
automated detection of end-diastolic and end-systolic contours with manual
correction if necessary. All radionuclide angiograms were interpreted by the
same experienced investigator, who was not aware of the CT results.
Statistical Analysis
Statistical analysis was performed with commercially available software
(SAS Institute). Statistical significance was set at p < 0.05.
Results were expressed as mean, standard deviation, and interquartile range
for continuous variables and as frequencies and percentages for categorical
variables. Concordance between MDCT and equilibrium radionuclide
ventriculography was assessed with intraclass correlation coefficient (ICC)
for continuous variables. The scale used for interpretation of this
concordance was that described by Fleiss
[24]. A concordance value of
0.6-0.8 was rated moderate and a value > 0.80, good. A Bland-Altman
[25] diagram was used for
further examination of concordance between the two techniques. Mean absolute
percentage error (MAPE) also was calculated. The 95% CI for mean difference
was an estimate of the limits of concordance between the two methods. MAPE is
the absolute value of the difference between the two measurements over the
mean of the two measurements expressed as a percentage.
The kappa coefficient was used to assess agreement between qualitative variables. The scale used to describe the degree of agreement was that proposed by Landis and Koch [26]. A degree of concordance < 0.20 was rated poor; 0.21-0.40, fair; 0.41-0.60, moderate; 0.61-0.80, good; and 0.81-1.00, excellent. The primary end point of this study was the ICC between the RVEF values estimated with two examinations.
The number of subjects needed was calculated by the method described by Walter et al. [27]. For a minimum acceptable ICC of 0.8, it was estimated that 46 patients were needed for 80% power with significance at 5% for an estimated intraclass correlation of 0.9. Additional statistical analyses were performed as follows. Mean attenuation values in the various regions of the right ventricle were compared by repeated analyses of variance. For continuous parameters, the results of CT and scintigraphy were compared by paired Student's t test or paired Wilcoxon's test when the sample size was less than 30. Comparisons between groups were performed with the Mann-Whitney test.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Calculation of RVEF with MDCT
Segmentation of RV cavity—Clear delineation of the
endocardial contours was possible in 34 (69%) of the examinations. In all of
these cases, homogeneous enhancement of the right ventricle without streak
artifacts was present around the ventricular cavity (mean attenuation value
within right ventricle, 214.62 ± 39.50 H; range, 144-302 H), and no
motion artifacts related to irregular cardiac rhythm were present. Imprecise
endocardial delineation was found in 15 (31%) of the examinations. The
delineation process was rated as imprecise because a low level of enhancement
within the right ventricle led to suboptimal delineation of the endocardial
contours (n = 11). The mean attenuation value in the right ventricles
of these patients was significantly lower than that measured in the group of
34 patients with images graded as having easily delineated endocardial
contours (164.94 ± 30.29 H; range, 120-234 H) (Mann-Whitney test,
p < 0.001). The presence of artifacts was related to extrasystolic
beats (n = 4), myocardial or pericardial calcification (n =
2), and/or presence of a pacemaker (n = 1). In these 15 cases,
alteration of image quality did not hamper the RV segmentation process.
Table 1 summarizes the
frequency of identification of the anatomic borders between the right
ventricle and the right atrium and between the right ventricle and the
pulmonary trunk.
|
|
Comparison of MDCT and Equilibrium Radionuclide Ventriculography
Mean RVEF was 43.63% ± 8.95% (range, 13-61%) with CT and 44.20%
± 11.68 (range, 13-66%) with equilibrium radionuclide ventriculography.
No significant difference was found between the mean values obtained with
either technique (paired Student's t test, p = 0.57).
Concordance between the two methods was moderate (ICC = 0.77). The dispersion
of differences in RVEF assessed with MDCT and scintigraphy are illustrated
with a Bland-Altman plot in Figure
4. Only three values (6% of the data) were outside the limits. The
limit of concordance for MDCT ranged from -14.9 to 13.7. The percentage of
variability between two measurements expressed by MAPE was 12.1%.
In a subanalysis, the mean value of RVEF obtained with CT was compared with that obtained with scintigraphy in two subgroups of patients according to the precision of RV cavity segmentation. There was no significant difference between the mean values determined with CT and scintigraphy in evaluation of the 34 patients for whom delineation of the endocardial contours was rated as precise (mean value with CT, 45.26% ± 8.89%; mean value with radionuclide ventriculography, 46.15% ± 12.43%; paired Student's test, p = 0.49) and the 15 patients for whom delineation of the endocardial contours was rated as imprecise (mean value with CT, 39.93% ± 8.21%; mean value with radionuclide ventriculography, 39.8% ± 8.56%; p = 0.83, Wilcoxon's test). The ICC was 0.77 for the 34 examinations with clear delineation of endocardial contours and 0.68 for the other 15 examinations.
According to the ventriculoscintigraphic results, 34 patients had a normal
RVEF (
40%) (mean RVEF, 50.06% ± 8.19%; range, 40-66%) and 15
patients had a low RVEF (< 40%) (mean RVEF, 30.93% ± 6.13%; range,
13-38%). Thirty-two of the 34 patients with a normal RVEF on
ventriculoscintigraphy had a normal RVEF on CT (mean RVEF, 48.08% ±
4.88%; range, 40-61%). Twelve of the 15 patients with a low RVEF on
ventriculoscintigraphy had a low RVEF on CT (mean RVEF, 32.5% ± 6.83%;
range, 13-39%). Thus there was a total of 44 (90%) concordant results between
the two techniques. Agreement between the two methods was good (
=
0.75) in a comparison of the subgroups of patients with normal and reduced
RVEF.
Radiation Dose
The mean dose-length product of functional scans was 427.77 ± 74.31
mGy · cm (range, 322-789 mGy · cm). The estimated mean radiation
dose of functional scans was 7.27 mSv. The mean theoretic dose for
ventriculoscintigraphy was 5 mSv. The mean dose-length product of diagnostic
scans was 257.45 ± 52.20 mGy · cm (range, 93-374 mGy ·
cm), leading to an estimated radiation dose of 4.37 mSv. The mean effective
dose for both chest and heart examinations was 11.64 mSv.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We investigated a CT protocol aimed at providing morphologic information on the underlying respiratory disease as well as functional information about the right ventricle during the same imaging session. To our knowledge, this investigation was the first aimed at integrating functional information into a diagnostic scan; previous investigations of RV function with MDCT focused on cardiac examinations [29, 30]. Using 16-MDCT technology, we designed a two-phase protocol. We started with non-ECG-gated acquisition over the entire thorax and followed with ECG-gated examination of the cardiac cavities with biphasic administration of contrast medium. Whereas the first phase was similar to a standard CT angiographic examination of the chest, the second phase was designed to allow acquisition of systolic and diastolic short-axis images of the right ventricle with a level of enhancement compatible with precise delineation of the endocardial contours. The time needed to obtain RVEF measurements varied between 11 and 15 minutes, including two time-consuming steps: creation of diastolic and systolic short-axis images of the cardiac cavities followed by RV segmentation in both cardiac phases, which is in the range reported for CT and MR volumetric left ventricular measurements [22]. The introduction of automated tools for generation of short-axis images of the cardiac cavities at various phases of the cardiac cycle is expected to shorten part of the procedure. With regard to the reconstruction parameters selected in our study, we decided to generate 3-mm-thick contiguous short-axis images of the cardiac cavities. This section thickness, thinner than that reported for CT and MR in evaluation of the left ventricle [22], was chosen to enhance sharpness of the RV endocardial contours and to facilitate precise recognition of the valvular borders of the right ventricle, a well-known potential source of error in scintigraphy [16]. Inclusion of the RV trabeculations and RV outflow tract led to overestimation of RV end-diastolic and end-systolic volumes but did not affect the overall estimate of RVEF.
Electron beam CT was the first imaging technique to be compared with MRI for assessment of RV volume and global ventricular function, and it had good agreement with a close correlation and acceptable interobserver variability [31]. Lembcke et al. [29] were the first authors to report their experience with 8- and 16-MDCT performed to measure RV dimensions and function. They conducted their study with a population of 25 patients undergoing cardiac surgery. Those authors found that compared with MRI, MDCT was an accurate and reliable noninvasive technique for evaluating RV measurements. In that study, volumetric evaluation was obtained on a stack of 5-mm-thick axial images with exclusion of the papillary muscles and trabeculae of the right ventricle. Despite different acquisition times for MDCT and MRI, the investigators found good agreement between the techniques, leading them to suggest that differences in temporal resolution between the two techniques are less important than systematic and random measurement errors resulting from the geometry of the right ventricle. Similar conclusions were recently been drawn by Koch et al. [30], who investigated assessment of RV function with 16-MDCT by using two software tools to compare CT with MRI in a population of 19 patients. At a gantry rotation time of 420 milliseconds, the temporal resolution achieved with CT in the study varied between 105 and 210 milliseconds, compared with 40-50 milliseconds for MRI [29]. Advances in MDCT technology, especially the introduction of 64-MDCT scanners, offer further improvement in this concept of integration of structure and function during CT examinations of the chest, which are first-line imaging tools for most patients with respiratory disease. With the latter technology, one can to obtain two sets of information from a single data set, but this capability requires further confirmation.
Assessing RV function with CT raises practical questions regarding radiation dose. In the present study, the mean of 7.27 mSv for the functional scan was higher than the mean theoretic dose delivered in radionuclide ventriculography, that is, 5 mSv [32]. However, CT provides not only functional but also morphologic information at the level of the cardiac cavities. In this preliminary investigation, we did not use dose-modulation software, which can be used efficiently in such examinations. There were several limitations to this study. First, we compared CT not with MRI, the accepted standard for assessment of RV function, but with radionuclide ventriculography, a method with limitations when applied to the right ventricle, especially in patients with abnormal RV shape and size [15]. However, this method has been shown to have adequate accuracy for routine clinical use [16, 18, 33]. Second, we did not evaluate intraindividual or interindividual variability of cardiac function measurements on CT scans. Because the study was aimed at assessing the accuracy of CT versus a reference method, we decided to focus on the consensus reading of CT scans and to compare these results with those of radionuclide ventriculography. This important practical issue is specifically addressed in a separate study [34]. Third, our analysis was based on a study group of 49 patients, a limitation directly linked to the lack of wide availability of equilibrium scintigraphy for RVEF measurement at our institution. Our conclusions need to be validated in a larger population.
In conclusion, this study shows the accuracy of RVEF measurements during CT angiographic examination of the chest performed with 16-MDCT equipment in patients with underlying respiratory disease.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Schroeder, S. Achenbach, F. Bengel, C. Burgstahler, F. Cademartiri, P. de Feyter, R. George, P. Kaufmann, A. F. Kopp, J. Knuuti, et al. Cardiac computed tomography: indications, applications, limitations, and training requirements: Report of a Writing Group deployed by the Working Group Nuclear Cardiology and Cardiac CT of the European Society of Cardiology and the European Council of Nuclear Cardiology Eur. Heart J., February 2, 2008; 29(4): 531 - 556. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
