Editorial Type:
Diagnostic Imaging

Comparison of multi-detector row computed tomography with echocardiography for assessment of left ventricular function in healthy dogs

Christiane R. Henjes Small Animal Clinic, Faculty of Veterinary Medicine, University of Veterinary Medicine Hannover, 30559 Hannover, Germany.

Search for other papers by Christiane R. Henjes in
Current site
Google Scholar
PubMed Close
,
Stephan Hungerbühler Small Animal Clinic, Faculty of Veterinary Medicine, University of Veterinary Medicine Hannover, 30559 Hannover, Germany.

Search for other papers by Stephan Hungerbühler in
Current site
Google Scholar
PubMed Close
 DVM
,
Iwona B. Bojarski Small Animal Clinic, Faculty of Veterinary Medicine, University of Veterinary Medicine Hannover, 30559 Hannover, Germany.

Search for other papers by Iwona B. Bojarski in
Current site
Google Scholar
PubMed Close
,
Ingo Nolte Small Animal Clinic, Faculty of Veterinary Medicine, University of Veterinary Medicine Hannover, 30559 Hannover, Germany.

Search for other papers by Ingo Nolte in
Current site
Google Scholar
PubMed Close
 DVM, PhD
, and
Patrick Wefstaedt Small Animal Clinic, Faculty of Veterinary Medicine, University of Veterinary Medicine Hannover, 30559 Hannover, Germany.

Search for other papers by Patrick Wefstaedt in
Current site
Google Scholar
PubMed Close
 DVM
DOI:
https://doi.org/10.2460/ajvr.73.3.393
Volume/Issue:
Volume 73: Issue 3
Online Publication Date:
01 Mar 2012
Full access
Sign In Reprints and Permissions

Abstract

Objective—To evaluate the use of retrospectively ECG-gated, contrast-enhanced, multi-detector row computed tomography (MDCT) for assessment of left ventricular function in dogs and to compare the results with those obtained by use of 2-D and M-mode echocardiographc techniques.

Animals—10 healthy Beagles.

Procedures—Dogs underwent MDCT (performed by use of a 64-detector row CT system) and echocardiography under general anesthesia. Left ventricular end-systolic volume (ESV), end-diastolic volume (EDV), and ejection fraction (EF) were determined in MDCT-generated multiplanar reformatted images by use of Simpson and biplane area-length calculation methods. Results were compared with left ventricular ESV, EDV, and EF determined in echocardiographc images by use of Teichholz and bullet method calculations. Results were evaluated via Deming regression analysis and Pearson correlation tests. Bland-Altman analysis was used to assess limits of agreement and systematic errors between the 2 methods.

Results—Mean values for EDV and ESV determined by use of MDCT were highly correlated with those determined by use of echocardiography, regardless of the calculation methods compared (r = 0.91 to 0.96); volumes determined by use of MDCT appeared to be higher than those determined by use of echocardiography, although most differences were nonsignificant. Mean EF determined by use of MDCT with the Simpson calculation method was highly correlated with that determined by use of echocardiography with bullet method calculations (r = 0.90).

Conclusions and Clinical Relevance—Results suggested that assessment of left ventricular volume and function in dogs is feasible with MDCT. To estimate left ventricular EF with MDCT. use of the Simpson calculation method is advised.

Abstract

Objective—To evaluate the use of retrospectively ECG-gated, contrast-enhanced, multi-detector row computed tomography (MDCT) for assessment of left ventricular function in dogs and to compare the results with those obtained by use of 2-D and M-mode echocardiographc techniques.

Animals—10 healthy Beagles.

Procedures—Dogs underwent MDCT (performed by use of a 64-detector row CT system) and echocardiography under general anesthesia. Left ventricular end-systolic volume (ESV), end-diastolic volume (EDV), and ejection fraction (EF) were determined in MDCT-generated multiplanar reformatted images by use of Simpson and biplane area-length calculation methods. Results were compared with left ventricular ESV, EDV, and EF determined in echocardiographc images by use of Teichholz and bullet method calculations. Results were evaluated via Deming regression analysis and Pearson correlation tests. Bland-Altman analysis was used to assess limits of agreement and systematic errors between the 2 methods.

Results—Mean values for EDV and ESV determined by use of MDCT were highly correlated with those determined by use of echocardiography, regardless of the calculation methods compared (r = 0.91 to 0.96); volumes determined by use of MDCT appeared to be higher than those determined by use of echocardiography, although most differences were nonsignificant. Mean EF determined by use of MDCT with the Simpson calculation method was highly correlated with that determined by use of echocardiography with bullet method calculations (r = 0.90).

Conclusions and Clinical Relevance—Results suggested that assessment of left ventricular volume and function in dogs is feasible with MDCT. To estimate left ventricular EF with MDCT. use of the Simpson calculation method is advised.

Analysis of ventricular function is used to investigate hemodynamics and to help determine prognosis in veterinary patients with various cardiovascular diseases including left ventricular hypertrophy, dilated cardiomyopathy,1 and valvular heart disease.2 Ventricular volume is the basis of cardiac function evaluation. Many cardiac indices such as EF are derived from ventricular volume. Other indices, such as ESV and EDV, require previous determination of ventricular volume.

To date, cardiac volume measurement in veterinary medicine has relied on 2-D modalities, such as cineangiocardiography,3,4 equilibrium radionuclide ventriculography,5–7 or echocardiography.8–10 All of these techniques require geometric modeling because 3-D volumes have to be calculated from 2-D measurements. In contrast, MDCT and MRI of the heart are true 3-D imaging modalities. Magnetic resonance imaging is the accepted gold standard method to measure left ventricular volume and function in humans. In the past, the use of MRI for ventricular volume determination in dogs was studied as well.11 However, in veterinary medicine, MRI of the heart is not commonly performed because extensive training is required for acquisition and postprocessing of the images. Furthermore, image acquisition during MRI is time-consuming because every MRI sequence must be matched to the subject in terms of heart rate and reasons for examination; there-fore, it requires prolonged anesthesia in comparison with MDCT examination.

Imaging of the moving heart requires a high temporal resolution to achieve an artifact-free display of myocardial contraction and chamber extension throughout the cardiac cycle. Application of CT to cardiac imaging has long been limited by insufficient temporal resolution because of slow gantry rotation and long total acquisition times that result from slow volume coverage with single-slice imaging. Recent technical advances in MDCT techniques provide new opportunities for cardiac imaging. The introduction of MDCT systems with subsecond rotation times has remarkably improved spatial and temporal resolution and has made MDCT a useful tool for evaluation of cardiac function12–18 in humans. Acquisition of data sets with isotropic voxels by use of a 64-slice MDCT scanner allows 3-D reconstruction of images and enables the user to make measurements in any desired plane, which should improve accuracy.

Initially, in human medicine, most attention was directed toward the validation of MDCT as a noninvasive method to acquire images of the coronary arterial tree.19 Computed tomography was proven to be useful for the evaluation of arterial plaque density in an in vitro model,20 as well as in preliminary studies of myocardial perfusion21 and morphology22,23 and calcification23 of cardiac valves in human patients. In MDCT of the heart, the image is continuously acquired throughout the entire cardiac cycle by use of a helical scanning technique. Simultaneously obtained ECG traces are retrospectively used to reconstruct images during a specific part of the cardiac cycle within the R-R interval.

Data obtained from contrast-enhanced MDCT of the heart provide information about cardiac morphology in addition to variables used to assess global left ventricular function, including ESV, EDV, and EF. Several techniques have been established in human medicine to calculate left ventricular volumes by use of MDCT data. In the Simpson calculation method, volume calculations are performed on the basis of measurements of the left ventricular cavity in contiguous MPR short-axis images of the heart; in contrast, the biplane area-length method relies on measurements obtained in long-axis MPR images perpendicular to the short axis.12,13 To date, only a few studies24–26 have described the use of MDCT. to obtain images of the heart in nonhuman animals, and investigations of the utility of MDCT for assessment of cardiac function in dogs are lacking.

Echocardiography is routinely used in veterinary medicine for cardiac imaging because of its noninvasiveness, accessibility, and low cost. The bullet and Teichholz methods can be used to calculate ventricular volumes in 2-D and M-mode echocardiographic images, respectively.8,10 Ultrasonography with bullet method calculations has been validated against cineangiocardiography for determination of left ventricular stroke volumes in dogs,4 and its use in this species has been further evaluated in other studies.8,27 Nevertheless, echocardiography has some limitations; for example, imaging from some angles can result in foreshortened views of the left ventricle, and different projections cannot be acquired during a single cardiac cycle. Multi-detector row CT may provide a suitable means of cardiac imaging without such limitations.

The purpose of the study reported here was to evaluate the use of retrospectively ECG-gated, contrast-enhanced MDCT for the assessment of global left ventricular function in dogs and to compare the results with those obtained by use of 2-D and M-mode echocardiographic techniques.

Materials and Methods

Animals—Ten healthy Beagles (4 females and 6 males; mean ± SD weight, 19.49 ± 2.30 kg [range, 17.2 to 22.7 kg]; mean ± SD age, 6.8 ± 1.08 years [range, 6 to 8 years]) were used in the study. All dogs were determined to be free of cardiovascular disease on the basis of results of physical examination and Doppler echocardiography. The dogs were owned by the University of Veterinary Medicine Hannover, and the study was approved by the Ethical Committee of the Lower Saxony State Office for Consumer Protection and Food Safety.

Anesthetic protocol for imaging studies—Anesthesia was induced with diazepama (0.5 mg/kg) and levomethadoneb (0.2 mg/kg) administered IV. Then, propofolc was administered to effect until a plane of anesthesia adequate to perform endotracheal intubation was achieved. Anesthesia was maintained with isofluraned (end-tidal concentration, 1.3%) in a gas mixture of 50% oxygen in air. Dogs were mechanically ventilated by use of a respirator.e The respiratory rate was set at 8 breaths/min, and tidal volume was adjusted to maintain end-tidal Paco2 between 40 and 45 mm Hg; isoflurane concentration and oxygen saturation were continuously measured with a calibrated monitor.f Heart rate was monitored by means of ECG, and systolic, diastolic, and mean arterial blood pressures were recorded every 5 minutes. Throughout the procedure, saline (0.9% NaCl) solution was administered (5 mL/kg/h, IV). Data acquisition was started 30 minutes after induction of anesthesia to ensure stability of anesthetic conditions. The MDCT examination was performed, followed by echocardiographic imaging. The total time from induction of anesthesia to the end of the imaging procedures was approximately 110 minutes for each dog. All dogs were monitored until recovery from anesthesia.

MDCT protocol and image acquisition—Contrast-enhanced MDCT examinations were performed by use of a 64-detector row CT system.g Dogs were placed in dorsal recumbency, ECG leads were attached to the forepaws and the left hind paw, and ECG tracings were recorded throughout spiral MDCT examination. Apnea was induced by means of manual hyperventilation (2 to 3 breaths) and was maintained for a maximum of 30 seconds. All scans were obtained during apnea with the following settings: collimation, 64 × 0.625 mm; table pitch, 0.20; tube voltage, 120 kV; tube current, 400 mA; and tube rotation time, 0.4 seconds. The field of view was adapted individually for each dog. Contrast mediumh (3 mL/kg, IV) was administered via a peripheral vein at a flow rate of 3 mL/s by use of a power injector.i The delay between injection of contrast medium and start of the spiral scan was determined with an automated bolus-tracking technique.28 For automatic detection of contrast medium arrival, the ascending aorta was selected as the region of interest, and the threshold for triggering the scan was set at 110 Hounsfield units. As soon as the threshold was exceeded, the MDCT scan was performed automatically.

Image reconstruction—Image series were reconstructed in increments of 5% of the R-R interval at multiple phases throughout the cardiac cycle. For image reconstruction, a multisegmental reconstruction algorithm, which is based on data interpolation between ≥ 2 consecutive cardiac cycles, was used. The matrix used was 512 × 512 pixels. Window settings were adapted individually. For further evaluation, MPR images were created according to the anatomic long axis and short axis of the left ventricle, rather than the CT transverse acquisition plane. For this purpose, the data were imported into commercially available software.j

In the planar view of the software, 3 planes of the heart oriented at 90° angles to each other (vertical long axis, horizontal long axis, and transverse short axis) were displayed. Although the software allows automatic generation of the cardiac axes, manual correction of these images was required. At first, a reformatted short-axis image was created in which the short-axis plane was parallel to the plane of the mitral valve. For further image reformation, a short-axis plane at midpapillary level was selected because detection of papillary muscles simplified orientation. In this short-axis view, the vertical long-axis plane was tilted until it dissected the left ventricle connecting the edges of the intraluminal aspects of the papillary muscles (Figure 1). The horizontal long-axis plane was automatically directed perpendicular to the vertical long-axis plane.

Figure 1—
Figure 1—

Representative MDCT-generated MPR images of the heart in a healthy adult Beagle. Images were obtained by use of retrospectively ECG-gated, contrast-enhanced, 64-detector row CT in a study to evaluate use of this method for assessment of left ventricular function in 10 dogs and to compare the results with respective values obtained by use of echocardiography. A—Shortaxis plane view at midpapillary level showing relative positions of the vertical (white bar) and horizontal (dashed bar) long-axis planes. B—Horizontal long-axis plane (4-chamber) view showing relative positions of the short-axis (gray bar) and vertical long-axis planes. C—Vertical long-axis plane (2-chamber) view showing relative positions of the short-axis and horizontal long-axis planes. Ao = Aorta. LA = Left atrium. LV = Left ventricle. MV = Mitral valve. RA = Right atrium. RV = Right ventricle.

Citation: American Journal of Veterinary Research 73, 3; 10.2460/ajvr.73.3.393

The reformatted images generated were approximations of the actual long-axis views, but displayed a shortened view of the left ventricle. In a final step, the vertical long-axis plane was adjusted parallel to the interventricular septum to connect the left ventricular apex and the center of the mitral orifice, displaying a 2-chamber view. The horizontal long-axis plane was oriented perpendicular to the vertical axis along a line from the apex of the left ventricle to the center of the mitral orifice, displaying a 4-chamber view (Figure 1).29,30

For global left ventricular function assessment, only end-diastolic and end-systolic phase images were required. The MPR images depicting end-diastole and end-systole were defined as those that had maximal and minimal left ventricular dimensions, respectively (Figure 2). Generation of MPR images according to anatomic planes of the heart, selection of end-diastole and end-systole images, and data analysis were performed by 1 experienced operator (CRH).

Figure 2—
Figure 2—

Representative MDCT-generated short-axis MPR images of the heart depicting end-diastolic (A) and end-systolic (B) phases of the cardiac cycle. See Figure 1 for key.

Citation: American Journal of Veterinary Research 73, 3; 10.2460/ajvr.73.3.393

MDCT data—Left ventricular volume was calculated by use of the Simpson calculation method,14 in which short-axis MPR images were used, and the biplane area-length method,12 in which vertical (2-chamber) and horizontal (4-chamber) long-axis MPR images were used. Left ventricular EDV and ESV measurements were obtained from each image set. Left ventricular EF was calculated by use of the following formula: (EDV − ESV)/EDV × 100%. All measurements were calculated 3 times; means of the 3 calculations were used for further statistical analysis.

Simpson calculation method—Multiplanar reformatted images created in short-axis orientation (slice thickness, 5 mm; no intersection gap) including the entire left ventricle from base to apex were used to obtain cross-sectional measurements (Figure 3). The most basal slice distal to the atrioventricular ring (ie, anulus fibrosus sinister cordis) and surrounded by at least 50% myocardium was defined as left ventricular base. The most apical slice with a visible lumen was defined as the apex of the left ventricle. Endocardial borders were traced manually in end-systole and end-diastole short-axis images.k Papillary muscles and trabeculations were considered to be part of the ventricular cavity for better comparability with echocardiographic images. The left ventricular volume for each dog at end-systole and end-diastole was calculated by use of the following equation: ∑AN × S, where AN represents all measured cross-sectional areas and S is section thickness.

Figure 3—
Figure 3—

Representative MDCT-generated MPR images of the heart used for Simpson method calculations of left ventricular volume. Images were obtained for each dog at endsystole and at end-diastole, endocardial borders were manually traced on each image, and cross-sectional areas were used to determine left ventricular volume. A—Images of the heart at end-diastole created in short-axis orientation parallel to the mitral valve. Images are shown progressively from the most apical (upper left) to the most basal (bottom right) slices. B—Representative horizontal long-axis view (top) and vertical long-axis view (bottom) MPR images showing apical and basal limits of the region from which short-axis images were generated. R = Right.

Citation: American Journal of Veterinary Research 73, 3; 10.2460/ajvr.73.3.393

Biplane area-length method—Integrated computer software was used to trace the interior of the left ventricle manually on previously created end-dia-stolic and end-systolic vertical and horizontal long-axis MPR images (Figure 4). Long-axis length was calculated from the left ventricular apex to the level of the mitral valve. The resulting area was provided automatically by the software. Area values as well as long-axis lengths of vertical and horizontal long-axis images were used by integrated software to calculate left ventricular volume automatically, according to the following equation: 8/(3 × ∏) × Area of vertical long-axis plane × Area of horizontal long-axis plane/LAL, where LAL is the length of the shorter of the 2 long axes in 2 dimensions.

Figure 4—
Figure 4—

Representative MDCT-generated horizontal (A and B) and vertical (C and D) long-axis MPR images of the heart at end-diastole (A and C) and end-systole (B and D). Integrated computer software was used to manually trace the endocardial border of the left ventricle and to calculate left ventricular volumes by use of the biplane area-length method. The diagonal lines represent long-axis length. HRA = High right atrium. LHA = Left high atrium. See Figure 1 for remainder of key.

Citation: American Journal of Veterinary Research 73, 3; 10.2460/ajvr.73.3.393

Echocardiography—Examination was performed in right lateral recumbency by use of a commercially available system.l A 7.0-MHz transducer was used to obtain images. The images were saved in cine-loop format triggered by ECG (4 to 6 cardiac cycles). Frames with the largest and smallest ventricular cavities were selected via visual assessment as the end-diastolic and end-systolic images, respectively. All measurements were made by 1 experienced operator (SH) on images of 3 consecutive heartbeats; the mean of these measurements was used for further statistical analysis.

Bullet method—Images were generated from the right parasternal short-axis view with probe placement at the right fourth or fifth intercostal space. The cross-sectional area of the endocardial tissue border of the left ventricle was traced on echocardiographic images in short-axis orientation at the level of chordae ten-dineae.6 Left ventricular length was measured on the right parasternal long-axis 5-chamber view, obtained to maximize the length of the left ventricle as the distance from the apex to the mitral-aortic junction.6,8,9 Volume at end-systole and end-diastole was calculated according to the following equation: 5/6 × A × L, where A is cross-sectional area of the left ventricle and L is left ventricular length.8

Teichholz method—Left ventricular measurements were obtained from the right parasternal long-axis view by use of 2-D-guided M-mode echocardiography. Left ventricular systolic and diastolic diameters were measured according to the leading-edge-to-leading-edge method.31 The EDV and ESV were calculated by use of the Teichholz formula10 as follows: V= (7 × LVd3)/(2.4 + LVd), where V is left ventricular volume and LVd is the measured left ventricular diameter at end-diastole or end-systole, respectively.

Statistical analysis—Heart rates during MDCT and during echocardiographic examination were compared by use of a paired Student t test after normal distribution of the data was confirmed by means of the Kolmogorov-Smirnov test. Left ventricular EDV, ESV, and EF are expressed as median and mean ± SD. Differences in measurements obtained by means of MDCT and echocardiography were compared by use of a Kruskall-Wallis test followed by a Dunn test for multiple comparisons by use of commercially available software.m Values of P < 0.05 were considered significant. Correlation between volumes determined by use of MDCT and echocardiography was analyzed via Deming regression analysis and Pearson correlation tests. Bland-Altman analysis was performed for each pair of values of left ventricular EDV, ESV, and EF to calculate limits of agreement and systematic errors between the MDCT and echocardiographic methods. For overall evaluation of MDCT performance, results obtained by use of ultrasonography with bullet-method calculations were considered reference values.

Results

Heart rate during image acquisition via MDCT ranged from 78 to 126 beats/min (median, 89.5 beats/min; mean ± SD, 94.6 ± 15.1 beats/min), whereas that during echocardiography ranged from 73 to 100 beats/min (median, 78.8 beats/min; mean ± SD, 79.6 ± 7.0 beats/min); heart rate was significantly (P < 0.01) lower during echocardiographic examination. The MDCT data acquisition was accomplished in < 10 seconds (range, 8.7 to 9.8 seconds), and data sets had good image quality for functional analysis in all dogs.

Median and mean ± SD left ventricular volumes determined by use of MDCT (with Simpson and biplane area-length calculation methods) and echocardiography (with Teichholz and bullet method calculations for M-mode and 2-D images, respectively) were summarized (Table 1). Volumes determined according to the 2 calculation methods for a given imaging modality were highly correlated (Table 2). Median values for EDV and ESV calculated by use of biplane area-length and Teichholz calculation methods were significantly (P < 0.05) different between the 2 imaging modalities. However, mean EDV and ESV values determined by use of 3 MDCT measurements were highly correlated with the respective values determined by use of echocardiography regardless of the calculation method used. Scatterplots of EDV, ESV, and EF calculated by use of the Simpson and bullet methods revealed an almost linear correlation between the 2 methods Figure 5).

Table 1—

Mean ± SD (median) values of variables* used to evaluate left ventricular function in 10 healthy adult Beagles.
 MDCTEchocardiography
VariableSimpson methodArea-length methodTeichholz methodBullet method
EDV (mL)57.72 ± 12.32 (52.95)61.50 ± 12.20 (55.33)a48.59 ± 11.42 (44.94)b52.17 ± 11.84 (49.29)
ESV (mL)35.20 ± 10.52 (33 30)35.64 ± 7.89 (34.17)a26.51 ± 9.72 (24.39)b28.68 ± 11.30 (25.02)
EF (%)39.50 ± 7.19 (37.10)42.03 ± 5.99 (41.67)46.22 ± 9.39 (44.60)45.85 ± 10.81 (44.28)

The ESV, EDV, and EF were determined in MDCT-generated MPR images by use of Simpson and biplane area-length calculation methods and in M-mode and 2-D echocardiographc images by use of Teichholz and bullet method calculations, respectively.

Within a row, median values with different superscripted letters differ significantly (Dunn's comparison test, P < 0.05).

Table 2—

Results of statistical comparisons between various methods used to calculate left ventricular volumes and EF* in the 10 dogs in Table 1.
 EDVESVEF
 CorrelationBland-Altman analysisCorrelationBland-Altman analysisCorrelationBland-Altman analysis
Methods comparedrP valueBiasSDrP valueBiasSDrP valueBiasSD
Bullet vs Simpson0.91< 0.015.555.170.96< 0.016.583.210.90< 0.01−6.355.39
Bullet vs area length0.92< 0.019.334.810.91< 0.016.965.330.620.06−3.828.50
Teichholz vs Simpson0.94< 0.019.134.230.95< 0.018.753.190.570.08−6.727.90
Teichholz vs area length0.96< 0.0112.913.590.91< 0.019.134.140.190.60−4.1910.15
Bullet vs Teichholz0.88< 0.01−3.585.610.92< 0.01−2.174.360.81< 0.010.366.34
Simpson vs area length0.99< 0.013.781.620.97< 0.010.383.540.85< 0.012.533.80

See Table 1 for key.

Figure 5—
Figure 5—

Scatter plots showing correlation between MDCT with Simpson method calculations and echocardiography with bullet method calculations for EDV (A), ESV (B), and EF (C) in 10 healthy adult Beagles. Results of echocardiography with bullet method calculations were considered reference values for evaluation of MDCT findings.

Citation: American Journal of Veterinary Research 73, 3; 10.2460/ajvr.73.3.393

Left ventricular EF determined by use of MDCT with the Simpson calculation method was highly correlated with that determined by use of echocardiography with bullet method calculations (r = 0.90; Figure 5; Table 2). However, correlations between EFs determined by use of the 2 imaging methods were not significant when other calculation methods were used.

The dispersion of differences of left ventricular EDV, ESV, and EF determined by use of the various MDCT and echocardiographic method calculations was evaluated by use of Bland-Altman analyses (Table 2). In general, Bland-Altman analyses revealed that left ventricular end-diastolic and ESV data appeared to range higher for MDCT measurements when plotted against the echocardiographic values, although the differences were not significant. In contrast, EF values appeared to be lower overall when measured by MDCT methods, compared with those determined by echocardiographic methods. Bland-Altman plots of MDCT with the Simpson calculation method and echocardiography with bullet method calculations exemplified these findings (Figure 6).

Figure 6—
Figure 6—

Bland-Altman plots of the differences between EDV (A), ESV (B), and EF (C) values determined by use of MDCT with Simpson method calculations and ultrasonography with bullet method calculations. The solid and dotted lines represent mean value of differences and limits of agreement (bias ± 1.96 SD), respectively.

Citation: American Journal of Veterinary Research 73, 3; 10.2460/ajvr.73.3.393

Discussion

In the present study, retrospectively ECG-gated, contrast-enhanced MDCT was used to determine left ventricular EDV, ESV, and EF in dogs, and results were compared with the same variables determined by use of 2-D and M-mode echocardiographic techniques. Simpson and biplane area-length calculation methods were used to calculate volumes in MDCT-generated MPR images, and bullet and Teichholz methods were used to calculate the corresponding volumes in 2-D and M-mode echocardiographic images, respectively. The bullet method is a simple 2-D planimetric echocardiographic technique for cardiac volume determination that has been evaluated in previous studies.4,8,27 Thus, although comparisons among all calculation methods were made, left ventricular volumes calculated by use of the bullet method were considered reference values in the study reported here.

Mean left ventricular EDV and ESV determined by use of MDCT with Simpson and biplane area-length calculation methods were highly correlated with those determined by use of echocardiography with bullet and Teichholz calculation methods, regardless of the calculation methods compared. Differences were nonsignificant, except for median EDV and ESV comparisons between MDCT with area-length method and echocardiography with Teichholz method calculations. Bland-Altman analysis appeared to indicate higher values for the MDCT measurements of left ventricular EDV and ESV when plotted against echocardiographic measurements; however, these differences were not significant. In previous studies13,32 in which MDCT of the heart was compared with echocardiography in humans, MDCT values were reportedly higher for variables used to assess cardiac function. Higher MDCT values could be ascribable to volume overestimation in MDCT-generated MPR images, volume underestimation in echocardiographic images, or both.

The Simpson calculation of left ventricular volume requires reformatting of acquired thin cross-sectional slices to short-axis MPR image slices that are 5 mm thick; this caused a slight blurring of endocardial margins in the present study. In addition, hyperdense pixels at the endocardial margin may have represented parts of the myocardium. This effect could explain possible overestimation of ventricular volumes. It is possible that further accuracy could be achieved for MDCT with the Simpson calculation method by reformatting data in the short-axis plane to thinner slices.33

In echocardiography, papillary muscles often appear to be indistinguishable from the left ventricular wall; thus, the decision of when to exclude or include parts of the muscle tissue can cause difficulties. For better comparability of measurements obtained by use of MDCT and echocardiography, papillary muscles and trabeculae were included in the ventricular cavity in the present study. This could also contribute to volume overestimation, especially when MDCT-generated images are used, because contrast medium fills the trabecular interspaces.

In MDCT with the Simpson calculation method, an accurate definition of the most basal slice is important because it has a substantial influence on left ventricular volume determinations. The basal slice contains the largest cross-sectional area of the MDCT MPR image stack, multiplied by the slice thickness of 5 mm. A potential error in volumetry might be caused by inaccurate slice selection. Thus, the most basal slice in end-diastolic and end-systolic MPR images was adjusted parallel to the plane of the mitral valve and covered the most basal portion of the left ventricle, distal to the atrioventricular ring.15

There are also limitations associated with echocardiographic quantitation of left ventricular volumes. These problems include foreshortening of views in which ventricular length is measured and difficulties in delineation of the endocardial border because of acoustic dropouts (loss of the echocardiographic signal) and thickness of the echocardiographically displayed tissue border, which can lead to volume underestimation by use of bullet method calculations.9 The most important drawback of echocardiographic volume calculation with the Teichholz method is that volume determination relies on 1-dimensional M-mode measurements. Thus, ventricular volume, which contains information about a 3-D structure, is extrapolated from a single dimension of the left ventricular long or short axis. The limitations of echocardiographic measurements are particularly evident if left ventricular dyssynergy is present.9,34

When left ventricular EF volumes were compared in the present study, values determined by use of echocardiography with bullet method calculations were significantly correlated with those obtained by use of MDCT with the Simpson calculation method, but not with those in which the biplane area-length calculation method was used. These results are in accordance with reports13 in human medicine in which Simpson and biplane area-length calculation methods were compared. Our results are not surprising; because Simpson calculation methods are based on area measurements in 10 to 14 MDCT slices, inaccuracies of endocardial border tracing in 1 slice would have little effect on the results. In contrast, inaccurate measurement in MDCT images in which biplane area-length method calculations are used might change the results considerably because these calculations rely on the analysis of only 2 perpendicular long-axis views. Thus, on the basis of geometric assumptions, MDCT with the Simpson calculation method should yield more accurate estimates of left ventricular volumes than MDCT with biplane area-length method calculations.

In contrast with the findings for EDV and ESV, EF volumes determined by use of MDCT were apparently lower than those determined by use of echocardiography and were likely underestimated. Theoretically, an insufficient temporal resolution results in systematic overestimation of left ventricular ESV and thus underestimation of calculated left ventricular EF. Therefore, an insufficient temporal resolution can result in an inability to capture the maximum systolic contraction. Inferior temporal resolution of MDCT in comparison with other modalities for left ventricular volume measurement is a commonly mentioned aspect in regard to its use in humans,35 and this information should not be neglected in veterinary studies. The temporal resolution of MDCT of the heart depends on 3 factors: image reconstruction algorithm, gantry rotation time, and heart rate.36,37 In the present study, data sets were reconstructed by use of a multisegment reconstruction algorithm based on data interpolation between ≥ 2 consecutive cardiac cycles. This approach is an effective means of improving temporal resolution. However, cardiac function and volume may differ from beat to beat, and sinus arrhythmias, which are common in dogs (but were not observed in the present study), may cause inaccuracies.

In the study reported here, challenges were encountered because commercial software programs for calculating ventricular volumes and EF, which were developed for use in humans, were used in dogs. Initial attempts to perform scans with dogs positioned in ventral recumbency (a more physiologic posture than dorsal recumbency) failed. In human medicine, MDCT of the heart is always performed with the patient in a supine position. Because of this, software programming for the Simpson calculation method relies on models that only accept images obtained from subjects positioned in this manner. The software also features semiautomatic contour detection of endocardial borders with a manual correction option, and although contrast enhancement of the ventricular chamber was good, extensive corrections in nearly all slices were necessary, and thus manual contour tracing was found to be easier but quite time-consuming. For volume determination according to the biplane area-length calculation method, MPR images were generated according to the horizontal and vertical long-axis measurements obtained from transverse slices. The software proposed these axes, but manual correction was necessary for each dog.

With respect to the determination of left ventricular performance by means of MDCT, potential sources of error must be considered (eg, the requirement of iodinated contrast media). There are only small differences in the Hounsfield-unit values of blood, myocardium, and cardiac valves; consequently, unenhanced MDCT images cannot be used to distinguish these structures clearly. Thus, delineating the ventricular wall from the blood requires injection of contrast media. The injection of contrast media adds to blood volume and might influence left ventricular volumes and EF in MDCT of the heart.

Another drawback of MDCT examinations is the need for general anesthesia and a short period of apnea to avoid motion artifacts during cardiac evaluation in animals. This approach may cause problems in patients that are potentially compromised because of heart disease. Furthermore, anesthesia may alter functional variables, and the measurements may not adequately reflect cardiac performance of the unanesthetized animal. In the present study, MDCT and echocardiographic examinations were performed during the same anesthetic period with comparable anesthetic conditions during both examinations. The MDCT and echocardiographic examinations were performed rapidly and sequentially to minimize effect of time under anesthesia. However, heart rates were significantly lower during echocardiographic examination, which might have been the result of the longer anesthesia period for these measurements.

Even though left ventricular volume measurement by use of MDCT has shortcomings, it is a simple and noninvasive technique, and benefits include high spatial resolution and short acquisition time in comparison with echocardiography and MRI. Additionally, a single MDCT image acquisition of the heart requiring approximately 10 seconds provides detailed anatomic information, which is difficult to obtain by use of other diagnostic techniques without need for extra acquisition time. The data set provides additional information about cardiac and pericardial structures as well as volume estimates and overcomes several problems of echocardiography such as left ventricular foreshortening or a need for different projections of the heart, which cannot be acquired during the same cardiac cycle. Further, MDCT of the heart is not affected by variables such as obesity or a small acoustic window. Because MDCT is a true volumetric modality, enlarged or grossly deformed hearts should not influence the accuracy of the measurements made in Simpson calculation method reconstructions because assumptions about left ventricular shape and geometry are avoided. A previous human study38 also showed high interobserver agreement among readers assessing MDCT images of the heart. Furthermore, the possibility of postprocessing is of great advantage. It is feasible to separate data acquisition and data evaluation, and these procedures can be performed by different individuals. Slice reformation for volume measurement or morphological evaluation can be repeated as many times as necessary, until satisfactory orientation is achieved. This should improve precision of the examination. The advantages of echocardiography for evaluation of the heart include lack of radiation exposure, ability to characterize blood flow, ability to provide real-time beat-to-beat analysis, and avoidance of anesthesia and iodinated contrast medium administration.

Limitations of the present study included the small number of dogs and their homogeneity in terms of weight, breed, age, and overall health. Larger studies are needed to confirm feasibility of MDCT for volume determination in dogs of various breeds and sizes. Because dogs in the study were considered healthy, no pathological left ventricular function and morphology regarding left ventricular function were found. Reliability of MDCT should be evaluated in dogs with impaired heart function. In the present study, MDCT and echocardiography were performed during the same anesthetic procedure, and MDCT was always performed first; thus, even though anesthetic conditions were stable, the effect of time under anesthesia could potentially have influenced results of the 2 examinations. We used echocardiography as a reference method. Even though echocardiography is often used as a standard for volume determination in dogs, it has shortcomings. Therefore, a study in which MDCT is directly compared with MRI as the gold standard method for functional cardiac imaging in dogs is desirable.

Our findings indicate that determination of left ventricular volumes is feasible with MDCT, whether Simpson or biplane area-length calculation methods are used, and that results are comparable to those obtained by use of echocardiography. For estimation of left ventricular EF, the MDCT with Simpson calculation method reconstruction should be preferred because values correlated more closely with those assessed by use of echocardiography.

ABBREVIATIONS

CT

Computed tomography

EDV

End-diastolic volume

EF

Ejection fraction

ESV

End-systolic volume

MDCT

Multi-detector row computed tomography

MPR

Multiplanar reformatted

MRI

Magnetic resonance imaging
a.

Diazepam-ratlopharm, Ratiopharm GmbH, Ulm, Germany.

b.

L-Polamivet, Intervet Deutschland GmbH, Unterschleissheim, Germany.

c.

Narcofol 10 mg/mL, CP-Pharma Handelsgesellschaft mbH, Burgdorf, Germany.

d.

Isoflurane CP, CP-Pharma Handelsgesellschaft mbH, Burgdorf, Germany.

e.

Dräger Ventilog2, Dräger Medical AG, Lübeck, Germany

f.

GE Datex-Ohmeda, GE Healthcare, Helsinki, Finland.

g.

Brilliance 64, Philips, Eindhoven, The Netherlands.

h.

Omnipaque-350, Bayer Vital, Leverkusen, Germany.

i.

Medrad Vistron CT Injection System, Medrad, Warrendale, Pa.

j.

Cardiac viewer, Philips, Eindhoven, The Netherlands.

k.

LV/RV Analysis, Philips, Eindhoven, The Netherlands.

l.

GE Vivid 7, GE Healthcare, Copenhagen, Denmark.

m.

GraphPad Prism, version 4.00 for Windows, GraphPad Software, San Diego, Calif.

References

  • 1.↑

    Borgarelli M, Santilli RA, Chiavegato D, et al. Prognostic indicators for dogs with dilated cardiomyopathy. J Vet Intern Med 2006; 20:104–110.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 2.↑

    Kittleson MD, Eyster GE, Kowlen GG, et al. Myocardial function in small dogs with chronic mitral regurgitation and severe congestive heart failure. J Am Vet Med Assoc 1984; 184:455–459.

    • Search Google Scholar
    • Export Citation
  • 3.

    Chapman CB, Baker O, Reynolds J, et al. Use of biplane cinefluorography for measurement of ventricular volume. Circulation 1958; 18:1105–1117.

  • 4.↑

    Gueret P, Meerbaum S, Zwehl W, et al. Two-dimensional echocardiographic assessment of left ventricular stroke volume: experimental correlation with thermodilution and cineangiography in normal and ischemic states. Cathet Cardiovasc Diagn 1981; 7:247–258.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5.

    van den Brom WE, Stokhof AA. Radionuclide angiography in the normal dog: equilibrium studies. Res Vet Sci 1989; 47:333–337.

  • 6.↑

    Sisson DD, Daniel GB, Twardock AR. Comparison of left ventricular ejection fractions determined in healthy anesthetized dogs by echocardiography and gated equilibrium radionuclide ventriculography. Am J Vet Res 1989; 50:1840–1847.

    • Search Google Scholar
    • Export Citation
  • 7.

    Hetyey C, Balogh L, Andócs G, et al. Comparison of echocardiography and gated equilibrium radionuclide ventriculography in the measurements of left ventricular systolic function parameters in healthy dogs. Acta Vet Hung 2002; 50:21–29.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 8.↑

    Wyatt HL, Heng MK, Meerbaum S, et al. Cross-sectional echocardiography. II. Analysis of mathematic models for quantifying volume of the formalin-fixed left ventricle. Circulation 1980; 61:1119–1125.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9.↑

    Gueret P, Meerbaum S, Wyatt HL, et al. Two-dimensional echocardiographic quantitation of left ventricular volumes and ejection fraction. Importance of accounting for dyssynergy in short-axis reconstruction models. Circulation 1980; 62:1308–1318.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10.↑

    Teichholz LE, Kreulen T, Hermann MV, et al. Problems in echocardiographic volume determinations: echocardiographic-angiographic correlations in the presence and absence of asynergy. Am J Cardiol 1976; 37:7–11.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 11.↑

    Dell'Italia LJ, Blackwell GG, Pearce DJ, et al. Assessment of ventricular volumes using cine magnetic resonance in the intact dog. A comparison of measurement methods. Invest Radiol 1994; 29:162–167.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 12.↑

    Juergens KU, Grude M, Fallenberg EM, et al. Using ECG-gated multidetector CT to evaluate global left ventricular myocardial function in patients with coronary artery disease. AJR Am J Roentgenol 2002; 179:1545–1550.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 13.↑

    Bansal D, Singh RM, Sakar M, et al. Assessment of left ventricular function: comparison of cardiac multidetector-row computed tomography with two-dimension standard echocardiography for assessment of left ventricular function. Int J Card Imaging 2008; 24:317–325.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 14.↑

    Hennemann MM, Schuijf JD, Jukema JW, et al. Assessment of global and regional left ventricular function and volumes with 64-slice MSCT: a comparison with 2D echocardiography. J Nucl Cardiol 2006; 13:480–487.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 15.↑

    Juergens KU, Grude M, Maintz D, et al. Multi-detector row CT of left ventricular function with dedicated analysis software versus MR imaging: initial experience. Radiology 2004; 230:403–410.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 16.

    Juergens KU & Fischbach R. Left ventricular function studied with MDCT. Eur Radiol 2006; 16:342–357.

  • 17.

    Heuschmid M, Rothfuss JK, Schroeder S, et al. Assessment of left ventricular myocardial function using 16-slice multidetectorrow computed tomography: comparison with magnetic resonance imaging and echocardiography. Eur Radiol 2006; 16:551–559.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18.

    Hosoi S, Mochizuki T, Miyagawa M, et al. Assessment of left ventricular volumes using multi-detector row computed tomography (MDCT): phantom and human studies. Radiat Med 2003; 21:62–67.

    • Search Google Scholar
    • Export Citation
  • 19.↑

    Achenbach S, Giesler T, Ropers D, et al. Detection of coronary artery stenoses by contrast-enhanced, retrospectively electrocardiographically-gated, multislice spiral computed tomography. Circulation 2001; 103:2535–2538.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 20.↑

    Schroeder S, Flohr T, Kopp AF, et al. Accuracy of density measurements within plaques located in artificial coronary arteries by x-ray multislice CT: results of a phantom study. J Comput Assist Tomogr 2001; 26:900–906.

    • Search Google Scholar
    • Export Citation
  • 21.↑

    Nikolaou K, Sanz J, Poon M, et al. Assessment of myocardial perfusion and viability from routine contrast-enhanced 16-de-tector-row computed tomography of the heart: preliminary results. Eur Radiol 2005; 15:864–871.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 22.

    Willmann JK, Kobza R, Roos JE, et al. ECG-gated multi-detector row CT for assessment of mitral valve disease: initial experience. Eur Radiol 2002; 12:2662–2669.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 23.↑

    Willmann JK, Weishaupt D, Lachat M, et al. Electrocardiographically gated multi-detector row CT for assessment of valvular morphology and calcification in aortic stenosis. Radiology 2002; 225:120–128.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 24.

    Berninger WH, Redington RW, Doherty P, et al. Gated cardiac scanning: canine studies. J Comput Assist Tomogr 1979; 3:155–163.

  • 25.

    Ringertz HG, Skiöldebrand CG, Refsum H, et al. A comparison between the information in gated and nongated cardiac CT images. J Comput Assist Tomogr 1982; 6:933–938.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 26.

    Mancini GBJ, Peck WW, Slutksy RA, et al. Use of computerized tomography to assess myocardial infarct size and ventricular function in dogs during acute coronary occlusion and reperfusion. Am J Cardiol 1984; 53:282–289.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 27.

    Wyatt HL, Haendchen RV, Meerbaum S, et al. Assessment of quantitative methods for two-dimensional echocardiography. Am J Cardiol 1983; 52:396–401.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 28.↑

    Cademartiri F, Nieman K, van der Lugt A, et al. Intravenous contrast material administration at 16-detector row helical CT coronary angiography: test bolus versus bolus tracking technique. Radiology 2004; 233:817–823.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 29.

    Standardization of cardiac tomographic imaging. From the Committee on Advanced Cardiac Imaging and Technology, Council on Clinical Cardiology, American Heart Association; Cardiovascular Imaging Committee, American College of Cardiology; and Board of Directors, Cardiovascular Council, Society of Nuclear Medicine. Circulation 1992; 86:338–339.

    • Search Google Scholar
    • Export Citation
  • 30.

    Cerqueira MD, Weissman NJ, Dilsizian V, et al. Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart: a statement for healthcare professionals from the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association. Circulation 2002; 105:539–542.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 31.↑

    Sahn DJ, DeMaria A, Kisslo J, et al. Recommendations regarding quantitation in M-mode echocardiography: results of a survey of echocardiographic measurements. Circulation 1987; 58:1072–1083.

    • Search Google Scholar
    • Export Citation
  • 32.

    Kim TH, Hur J, Kim SJ, et al. Two-phase reconstruction for the assessment of left ventricular volume and function using retrospective ECG-gated MDCT: comparison with echocardiography. AJR Am J Roentgenol 2005; 185:319–325.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 33.↑

    Cui W, Kondo T, Anno H, et al. The accuracy and optimal slice thickness of multislice helical computed tomography for right and left ventricular volume measurement. Chin Med J (Engl) 2004; 117:1283–1287.

    • Search Google Scholar
    • Export Citation
  • 34.

    Serres F, Chetboul V, Tissier R, et al. Comparison of 3 ultrasound methods for quantifying left ventricular systolic function: correlation with disease severity and prognostic value in dogs with mitral valve disease. J Vet Intern Med 2008; 22:566–577.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 35.↑

    Heuschmid M, Küttner A, Schröder S, et al. Left ventricular functional parameters using ECG-gated multidetector spiral CT in comparison with invasive ventriculography. Rofo 2003; 175:1349–1354.

    • Search Google Scholar
    • Export Citation
  • 36.

    Flohr T, Bruder H, Stierstorfer K, et al. New technical developments in multislice CT, part 2: sub-millimeter 16-slice scanning an increased gantry rotation speed for cardiac imaging. Rofo 2002; 174:1022–1027.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 37.

    Mahnken A, Hohl C, Suess C, et al. Influence of heart rate and temporal resolution on left-ventricular volumes in cardiac multislice spiral computed tomography. A phantom study. Invest Radiol 2006; 41:429–435.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 38.↑

    Boehm T, Alkadhi H, Roffi M, et al. Time-effectiveness, observer-dependence, and accuracy of measurements of left ventricular ejection fraction using 4-channel MDCT. Rofo 2004; 176:529–537.

    • Crossref
    • Search Google Scholar
    • Export Citation

Contributor Notes

Presented in abstract form at the 17th Annual InnLab Meeting, Berlin, February 2009.

Address correspondence to Dr. Wefstaedt (Patrick.wefstaedt@tiho-hannover.de).
Cover American Journal of Veterinary Research
American Journal of Veterinary Research
Publication Date:
01 Mar 2012
  • Figure 1—
    View in gallery
    Figure 1—

    Representative MDCT-generated MPR images of the heart in a healthy adult Beagle. Images were obtained by use of retrospectively ECG-gated, contrast-enhanced, 64-detector row CT in a study to evaluate use of this method for assessment of left ventricular function in 10 dogs and to compare the results with respective values obtained by use of echocardiography. A—Shortaxis plane view at midpapillary level showing relative positions of the vertical (white bar) and horizontal (dashed bar) long-axis planes. B—Horizontal long-axis plane (4-chamber) view showing relative positions of the short-axis (gray bar) and vertical long-axis planes. C—Vertical long-axis plane (2-chamber) view showing relative positions of the short-axis and horizontal long-axis planes. Ao = Aorta. LA = Left atrium. LV = Left ventricle. MV = Mitral valve. RA = Right atrium. RV = Right ventricle.

  • Figure 2—
    View in gallery
    Figure 2—

    Representative MDCT-generated short-axis MPR images of the heart depicting end-diastolic (A) and end-systolic (B) phases of the cardiac cycle. See Figure 1 for key.

  • Figure 3—
    View in gallery
    Figure 3—

    Representative MDCT-generated MPR images of the heart used for Simpson method calculations of left ventricular volume. Images were obtained for each dog at endsystole and at end-diastole, endocardial borders were manually traced on each image, and cross-sectional areas were used to determine left ventricular volume. A—Images of the heart at end-diastole created in short-axis orientation parallel to the mitral valve. Images are shown progressively from the most apical (upper left) to the most basal (bottom right) slices. B—Representative horizontal long-axis view (top) and vertical long-axis view (bottom) MPR images showing apical and basal limits of the region from which short-axis images were generated. R = Right.

  • Figure 4—
    View in gallery
    Figure 4—

    Representative MDCT-generated horizontal (A and B) and vertical (C and D) long-axis MPR images of the heart at end-diastole (A and C) and end-systole (B and D). Integrated computer software was used to manually trace the endocardial border of the left ventricle and to calculate left ventricular volumes by use of the biplane area-length method. The diagonal lines represent long-axis length. HRA = High right atrium. LHA = Left high atrium. See Figure 1 for remainder of key.

  • Figure 5—
    View in gallery
    Figure 5—

    Scatter plots showing correlation between MDCT with Simpson method calculations and echocardiography with bullet method calculations for EDV (A), ESV (B), and EF (C) in 10 healthy adult Beagles. Results of echocardiography with bullet method calculations were considered reference values for evaluation of MDCT findings.

  • Figure 6—
    View in gallery
    Figure 6—

    Bland-Altman plots of the differences between EDV (A), ESV (B), and EF (C) values determined by use of MDCT with Simpson method calculations and ultrasonography with bullet method calculations. The solid and dotted lines represent mean value of differences and limits of agreement (bias ± 1.96 SD), respectively.

Figure 1—

Representative MDCT-generated MPR images of the heart in a healthy adult Beagle. Images were obtained by use of retrospectively ECG-gated, contrast-enhanced, 64-detector row CT in a study to evaluate use of this method for assessment of left ventricular function in 10 dogs and to compare the results with respective values obtained by use of echocardiography. A—Shortaxis plane view at midpapillary level showing relative positions of the vertical (white bar) and horizontal (dashed bar) long-axis planes. B—Horizontal long-axis plane (4-chamber) view showing relative positions of the short-axis (gray bar) and vertical long-axis planes. C—Vertical long-axis plane (2-chamber) view showing relative positions of the short-axis and horizontal long-axis planes. Ao = Aorta. LA = Left atrium. LV = Left ventricle. MV = Mitral valve. RA = Right atrium. RV = Right ventricle.
Figure 2—

Representative MDCT-generated short-axis MPR images of the heart depicting end-diastolic (A) and end-systolic (B) phases of the cardiac cycle. See Figure 1 for key.
Figure 3—

Representative MDCT-generated MPR images of the heart used for Simpson method calculations of left ventricular volume. Images were obtained for each dog at endsystole and at end-diastole, endocardial borders were manually traced on each image, and cross-sectional areas were used to determine left ventricular volume. A—Images of the heart at end-diastole created in short-axis orientation parallel to the mitral valve. Images are shown progressively from the most apical (upper left) to the most basal (bottom right) slices. B—Representative horizontal long-axis view (top) and vertical long-axis view (bottom) MPR images showing apical and basal limits of the region from which short-axis images were generated. R = Right.
Figure 4—

Representative MDCT-generated horizontal (A and B) and vertical (C and D) long-axis MPR images of the heart at end-diastole (A and C) and end-systole (B and D). Integrated computer software was used to manually trace the endocardial border of the left ventricle and to calculate left ventricular volumes by use of the biplane area-length method. The diagonal lines represent long-axis length. HRA = High right atrium. LHA = Left high atrium. See Figure 1 for remainder of key.
Figure 5—

Scatter plots showing correlation between MDCT with Simpson method calculations and echocardiography with bullet method calculations for EDV (A), ESV (B), and EF (C) in 10 healthy adult Beagles. Results of echocardiography with bullet method calculations were considered reference values for evaluation of MDCT findings.
Figure 6—

Bland-Altman plots of the differences between EDV (A), ESV (B), and EF (C) values determined by use of MDCT with Simpson method calculations and ultrasonography with bullet method calculations. The solid and dotted lines represent mean value of differences and limits of agreement (bias ± 1.96 SD), respectively.
  • 1.

    Borgarelli M, Santilli RA, Chiavegato D, et al. Prognostic indicators for dogs with dilated cardiomyopathy. J Vet Intern Med 2006; 20:104–110.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 2.

    Kittleson MD, Eyster GE, Kowlen GG, et al. Myocardial function in small dogs with chronic mitral regurgitation and severe congestive heart failure. J Am Vet Med Assoc 1984; 184:455–459.

    • Search Google Scholar
    • Export Citation
  • 3.

    Chapman CB, Baker O, Reynolds J, et al. Use of biplane cinefluorography for measurement of ventricular volume. Circulation 1958; 18:1105–1117.

  • 4.

    Gueret P, Meerbaum S, Zwehl W, et al. Two-dimensional echocardiographic assessment of left ventricular stroke volume: experimental correlation with thermodilution and cineangiography in normal and ischemic states. Cathet Cardiovasc Diagn 1981; 7:247–258.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5.

    van den Brom WE, Stokhof AA. Radionuclide angiography in the normal dog: equilibrium studies. Res Vet Sci 1989; 47:333–337.

  • 6.

    Sisson DD, Daniel GB, Twardock AR. Comparison of left ventricular ejection fractions determined in healthy anesthetized dogs by echocardiography and gated equilibrium radionuclide ventriculography. Am J Vet Res 1989; 50:1840–1847.

    • Search Google Scholar
    • Export Citation
  • 7.

    Hetyey C, Balogh L, Andócs G, et al. Comparison of echocardiography and gated equilibrium radionuclide ventriculography in the measurements of left ventricular systolic function parameters in healthy dogs. Acta Vet Hung 2002; 50:21–29.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 8.

    Wyatt HL, Heng MK, Meerbaum S, et al. Cross-sectional echocardiography. II. Analysis of mathematic models for quantifying volume of the formalin-fixed left ventricle. Circulation 1980; 61:1119–1125.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9.

    Gueret P, Meerbaum S, Wyatt HL, et al. Two-dimensional echocardiographic quantitation of left ventricular volumes and ejection fraction. Importance of accounting for dyssynergy in short-axis reconstruction models. Circulation 1980; 62:1308–1318.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10.

    Teichholz LE, Kreulen T, Hermann MV, et al. Problems in echocardiographic volume determinations: echocardiographic-angiographic correlations in the presence and absence of asynergy. Am J Cardiol 1976; 37:7–11.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 11.

    Dell'Italia LJ, Blackwell GG, Pearce DJ, et al. Assessment of ventricular volumes using cine magnetic resonance in the intact dog. A comparison of measurement methods. Invest Radiol 1994; 29:162–167.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 12.

    Juergens KU, Grude M, Fallenberg EM, et al. Using ECG-gated multidetector CT to evaluate global left ventricular myocardial function in patients with coronary artery disease. AJR Am J Roentgenol 2002; 179:1545–1550.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 13.

    Bansal D, Singh RM, Sakar M, et al. Assessment of left ventricular function: comparison of cardiac multidetector-row computed tomography with two-dimension standard echocardiography for assessment of left ventricular function. Int J Card Imaging 2008; 24:317–325.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 14.

    Hennemann MM, Schuijf JD, Jukema JW, et al. Assessment of global and regional left ventricular function and volumes with 64-slice MSCT: a comparison with 2D echocardiography. J Nucl Cardiol 2006; 13:480–487.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 15.

    Juergens KU, Grude M, Maintz D, et al. Multi-detector row CT of left ventricular function with dedicated analysis software versus MR imaging: initial experience. Radiology 2004; 230:403–410.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 16.

    Juergens KU & Fischbach R. Left ventricular function studied with MDCT. Eur Radiol 2006; 16:342–357.

  • 17.

    Heuschmid M, Rothfuss JK, Schroeder S, et al. Assessment of left ventricular myocardial function using 16-slice multidetectorrow computed tomography: comparison with magnetic resonance imaging and echocardiography. Eur Radiol 2006; 16:551–559.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18.

    Hosoi S, Mochizuki T, Miyagawa M, et al. Assessment of left ventricular volumes using multi-detector row computed tomography (MDCT): phantom and human studies. Radiat Med 2003; 21:62–67.

    • Search Google Scholar
    • Export Citation
  • 19.

    Achenbach S, Giesler T, Ropers D, et al. Detection of coronary artery stenoses by contrast-enhanced, retrospectively electrocardiographically-gated, multislice spiral computed tomography. Circulation 2001; 103:2535–2538.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 20.

    Schroeder S, Flohr T, Kopp AF, et al. Accuracy of density measurements within plaques located in artificial coronary arteries by x-ray multislice CT: results of a phantom study. J Comput Assist Tomogr 2001; 26:900–906.

    • Search Google Scholar
    • Export Citation
  • 21.

    Nikolaou K, Sanz J, Poon M, et al. Assessment of myocardial perfusion and viability from routine contrast-enhanced 16-de-tector-row computed tomography of the heart: preliminary results. Eur Radiol 2005; 15:864–871.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 22.

    Willmann JK, Kobza R, Roos JE, et al. ECG-gated multi-detector row CT for assessment of mitral valve disease: initial experience. Eur Radiol 2002; 12:2662–2669.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 23.

    Willmann JK, Weishaupt D, Lachat M, et al. Electrocardiographically gated multi-detector row CT for assessment of valvular morphology and calcification in aortic stenosis. Radiology 2002; 225:120–128.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 24.

    Berninger WH, Redington RW, Doherty P, et al. Gated cardiac scanning: canine studies. J Comput Assist Tomogr 1979; 3:155–163.

  • 25.

    Ringertz HG, Skiöldebrand CG, Refsum H, et al. A comparison between the information in gated and nongated cardiac CT images. J Comput Assist Tomogr 1982; 6:933–938.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 26.

    Mancini GBJ, Peck WW, Slutksy RA, et al. Use of computerized tomography to assess myocardial infarct size and ventricular function in dogs during acute coronary occlusion and reperfusion. Am J Cardiol 1984; 53:282–289.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 27.

    Wyatt HL, Haendchen RV, Meerbaum S, et al. Assessment of quantitative methods for two-dimensional echocardiography. Am J Cardiol 1983; 52:396–401.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 28.

    Cademartiri F, Nieman K, van der Lugt A, et al. Intravenous contrast material administration at 16-detector row helical CT coronary angiography: test bolus versus bolus tracking technique. Radiology 2004; 233:817–823.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 29.

    Standardization of cardiac tomographic imaging. From the Committee on Advanced Cardiac Imaging and Technology, Council on Clinical Cardiology, American Heart Association; Cardiovascular Imaging Committee, American College of Cardiology; and Board of Directors, Cardiovascular Council, Society of Nuclear Medicine. Circulation 1992; 86:338–339.

    • Search Google Scholar
    • Export Citation
  • 30.

    Cerqueira MD, Weissman NJ, Dilsizian V, et al. Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart: a statement for healthcare professionals from the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association. Circulation 2002; 105:539–542.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 31.

    Sahn DJ, DeMaria A, Kisslo J, et al. Recommendations regarding quantitation in M-mode echocardiography: results of a survey of echocardiographic measurements. Circulation 1987; 58:1072–1083.

    • Search Google Scholar
    • Export Citation
  • 32.

    Kim TH, Hur J, Kim SJ, et al. Two-phase reconstruction for the assessment of left ventricular volume and function using retrospective ECG-gated MDCT: comparison with echocardiography. AJR Am J Roentgenol 2005; 185:319–325.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 33.

    Cui W, Kondo T, Anno H, et al. The accuracy and optimal slice thickness of multislice helical computed tomography for right and left ventricular volume measurement. Chin Med J (Engl) 2004; 117:1283–1287.

    • Search Google Scholar
    • Export Citation
  • 34.

    Serres F, Chetboul V, Tissier R, et al. Comparison of 3 ultrasound methods for quantifying left ventricular systolic function: correlation with disease severity and prognostic value in dogs with mitral valve disease. J Vet Intern Med 2008; 22:566–577.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 35.

    Heuschmid M, Küttner A, Schröder S, et al. Left ventricular functional parameters using ECG-gated multidetector spiral CT in comparison with invasive ventriculography. Rofo 2003; 175:1349–1354.

    • Search Google Scholar
    • Export Citation
  • 36.

    Flohr T, Bruder H, Stierstorfer K, et al. New technical developments in multislice CT, part 2: sub-millimeter 16-slice scanning an increased gantry rotation speed for cardiac imaging. Rofo 2002; 174:1022–1027.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 37.

    Mahnken A, Hohl C, Suess C, et al. Influence of heart rate and temporal resolution on left-ventricular volumes in cardiac multislice spiral computed tomography. A phantom study. Invest Radiol 2006; 41:429–435.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 38.

    Boehm T, Alkadhi H, Roffi M, et al. Time-effectiveness, observer-dependence, and accuracy of measurements of left ventricular ejection fraction using 4-channel MDCT. Rofo 2004; 176:529–537.

    • Crossref
    • Search Google Scholar
    • Export Citation

Advertisement