One of the most common applications of echocardiography in veterinary and human medicine is the assessment of LV size and function. Several common cardiac diseases of dogs, including myxomatous mitral valve disease and dilated cardiomyopathy, affect LV size and function, and calculated echocardiographic values are used as a tool in clinical decision making.1,2 Thus, accurate and quantitative evaluation of the LV is critical for the clinical diagnosis, treatment, and estimation of prognosis in patients with occult or clinical cardiac disease. The quantification of LV volumes and systolic function can be assessed with multiple imaging modalities. Technological advances in echocardiography and CT have expanded the ability of clinicians to assess LV volume and function, and each of these modalities has its own attributes and limitations.
Echocardiography is an excellent modality for the assessment of LV size and function because it is noninvasive, widely available, and safe. Historically, 1-D images obtained in M-mode were used to estimate LV volume by calculating the cube of linear measurements of LV width obtained during end diastole and end systole in the Teichholz formula.3 Two-dimensional echocardiography has replaced M-mode echocardiography in humans undergoing an echocardiographic evaluation,4 but M-mode echocardiography is still frequently used in veterinary medicine for the estimation of LV volumes. The American Society of Echocardiography recommends LV volumes be estimated via 3-D echocardiography or, if such images are suboptimal, from 2-D images by use of the modified Simpson MOD or ellipsoid method from single-plane or biplane images.4 The main limitation of 1-D and 2-D echocardiography for estimating LV volume is the geometric shape assumptions required to calculate a 3-D volume from 1-D or 2-D measurements. In addition to the pitfalls of geometric modeling, foreshortened LV views, erroneous detection of the endocardial border, images from separate cardiac cycles, and malaligned M-mode cursor placement also contribute to inaccuracies of echocardiographic LV volume measurements.4,5
The development of matrix-array transducers and RT3DE has advanced the ability to quantify LV volumes and function echocardiographically without geometric modeling. Semiautomated and fully automated software for detection of the endocardial border generate a 3-D cast of the LV cavity and calculate LV volumes throughout the cardiac cycle. Studies and meta-analyses in human cardiology have reported superior accuracy of RT3DE, compared with results for 2-D echocardiographic methods, for quantification of LV volume.6–8 Real-time 3-D echocardiography is currently not widely available in veterinary medicine; however, it has been used in studies to assess LV9–11 and left atrial12,13 volumes and function in dogs with and without myxomatous mitral valve disease. In addition to 3-D volumetric measurements, the pyramidal data collected with RT3DE can be used to simultaneously display the 4- and 2-chamber orthogonal LV views and enable RTBPE LV measurements by use of the modified Simpson MOD. One of the major obstacles to performing biplane rather than single-plane measurements in human and veterinary patients is the difficulty of obtaining the left apical 2-chamber view.4,14 Use of RTBPE overcomes this limitation by acquiring 3-D data of the entire LV and displaying 2 truly orthogonal views.
Advances in CT technology during the past 2 decades have substantially increased the utility of this imaging modality for cardiac assessment. Development of slip-ring technology, multiple detector rows, and ECG gating has improved spatial and temporal resolution to a degree that allows analysis of LV function. Additionally, the use of injectable contrast agents results in excellent detection of the endocardial border for LV chamber quantification with automated software. Although cardiac MRI is typically used as the criterion-referenced standard in human cardiology, comparisons of MDCT and MRI have revealed excellent correlations for human15–18 and veterinary19 studies. Cardiac MRI and MDCT both provide accurate measurements of LV volume and function,15–20 although the requirement for anesthesia and expensive imaging equipment limits their usefulness for routine clinical assessment of cardiac function in veterinary medicine.
The accuracy of RT3DE for assessment of LV volume and function has been evaluated rigorously in human medicine; however, to the authors’ knowledge, it has been investigated only once in veterinary medicine.11 The purpose of the study reported here was to examine healthy dogs and compare LV volume and function variables assessed by use of 1-D, 2-D, and 3-D echocardiographic methods with results for ECG-gated MDCT angiography, which was used as a criterion-referenced standard. Our hypothesis was that 3-D echocardiographic methods would yield a higher correlation with MDCT than would 1-D and 2-D echocardiographic methods.
Materials and Methods
Animals
Six healthy purpose-bred sexually intact female dogs were used in the study. There were 4 Beagles and 2 mixed-breed hound-type dogs. Median body weight was 11.8 kg (range, 10.2 to 29.0 kg), and median age was 2 years (range, 1 to 3 years). All dogs were considered to be healthy on the basis of results of a physical examination, CBC, biochemical panel, and 2-D Doppler echocardiography. An a priori power analysisa was performed by use of MRI and 3-D echocardiography EDV values of dogs11; that analysis yielded an estimated power of 0.83 with a sample size of 6 dogs. The study was approved by the Oregon State University Institutional Animal Care and Use Committee.
Anesthesia and infusion of esmolol
Dogs were anesthetized for echocardiography and MDCT. Each dog was premedicated with hydromorphone (0.1 mg/kg, IM), and anesthesia was induced with propofol (IV, to effect). Dogs were endotracheally intubated, and anesthesia was maintained by administration of isoflurane in oxygen and use of mechanical ventilation. A constant rate infusion of esmolol was administered to assist heart rate control for MDCT ECG gating. Ten minutes after induction of anesthesia, a loading dose of esmolol (0.25 mg/kg, IV) was administered, which was followed by a constant rate infusion of esmolol (100 μg/kg/min). Dogs were allowed to achieve stable anesthetic conditions. Data acquisition started 20 minutes after induction of anesthesia. The dogs were arbitrarily assigned to the order of echocardiography and MDCT in an alternating scheme to minimize the effects of anesthetic duration on ventricular function. Thus, 3 dogs completed MDCT followed by echocardiography and 3 dogs completed echocardiography followed by MDCT. From the time of positioning on the table to the end of the study, the time for acquisition of MDCT images was approximately 15 minutes, whereas the time for acquisition of echocardiography images was approximately 40 to 60 minutes. All dogs were monitored until fully recovered from anesthesia.
M-mode and 2-D echocardiography
Echocardiographic examinations were performed by a board-certified veterinary cardiologist (KFS). The investigator was not blinded with regard to the echocardiographic analysis, although all echocardiographic measurements were performed before MDCT to exclude influence of MDCT results on echocardiographic variables. Images were obtained from standard right parasternal short- and long-axis views and left apical imaging planes by use of an ultrasound unitb equipped with a 1- to 5-MHz matrix transducer. Still images and cine loops of 5 cardiac cycles were recorded and transferred to a workstationc for off-line analysis. Each measurement was performed in triplicate, and the mean was calculated and used for statistical analysis. The M-mode and 2-D echocardiographic images were obtained and measured as recommended in guidelines established by the American Society of Echocardiography.4 The M-mode echocardiography of the LV was performed from the right parasternal short-axis view at the level of the papillary muscles. Two-dimensional cine loops were obtained for the 4- and 2-chamber views from the left apical window and from the LV short-axis view at the level of the mitral valve and papillary muscles. We attempted to obtain optimal visualization of the LV apex from the apical window in all dogs to avoid apical foreshortening.
Teichholz measurement
The LV diameter was measured at end diastole and end systole by use of the leading edge-to-leading edge technique. End diastole was defined as the beginning of the QRS complex, and end systole was defined as the closest approximation of the septum and free wall. Values for EDV and ESV were calculated by use of the Teichholz equation,3 SV was calculated as EDV – ESV, and EF was calculated as ([EDV – ESV]/EDV) × 100.
MOD and area-length measurements
Single-plane MOD and area-length measurements of the LV were made from the left apical 4-chamber view at end diastole and end systole. Biplane MOD and area-length measurements were made from the 4- and 2-chamber views of the LV. Manual tracing of the LV endocardial border was performed from one mitral valve annulus to the other and included the papillary muscles within the LV cavity. The LV length was measured from the midpoint of the mitral valve annulus to the LV endocardial apex. Values for EDV and ESV were automatically calculated by the workstation software. The MOD calculations were made by use of the single-plane and biplane modified Simpson rule, and area-length calculations were performed by use of the single-plane and biplane ellipsoid formula as follows: V = ([5/6A2]/L), where V is LV volume, A is LV area, and L is LV length.
3-D echocardiography
A 3-D data set was acquired from the left apical window by use of 4 consecutive ECG-triggered subvolumes integrated into 1 pyramidal full volume. Full volumes were analyzed off-line by use of software.d,e For RTBPE measurements, 4-chamber, 2-chamber, short-axis, and 3-D views were simultaneously displayed in a quad screen by use of 3-D quantification software (Figure 1). The 4- and 2-chamber views were manually adjusted to optimize LV length and avoid foreshortening. Manual tracing of the LV endocardial border was performed at end diastole and end systole, and the EDV and ESV were calculated by use of the modified Simpson MOD. Real-time 3D volumes were measured with 3-D advanced quantification software at end diastole and end systole (Figure 2). End diastole was defined as the first frame in which there was mitral valve closure, and end systole was defined as the last frame before mitral valve opening. Manual identification of 5 reference points was performed, 4 around the mitral valve annulus and 1 at the LV apex, on the 4- and 2-chamber views at end diastole and end systole. The software then automatically detected the endocardial border to create a cast of the LV throughout the cardiac cycle. Manual adjustments were made for the border detection in each frame to optimize the computer-generated LV cast. Values for EDV and ESV were calculated automatically by the software.
The RTBPE images of the left apical 4-chamber (A) and 2-chamber (B) views, short- axis view (C), and 3-D pyramid view (D) from the same cardiac cycle for the 3-D full volume LV data set of a clinically normal dog. Manual tracing of the LV endocardial border on the 4- and 2-chamber views allows for a real-time biplane modified Simpson MOD calculation of LV, EDV, and ESV. The red, blue, and green lines represent the planes that are viewed in the corresponding green, red, and blue box within panels A, B, and C, respectively. In panel C, these planes illustrate the orthogonal relationship of the 4- and 2-chamber views, and their location within the 3-D volume view is depicted in panel D.
Citation: American Journal of Veterinary Research 77, 11; 10.2460/ajvr.77.11.1211
The RT3DE images of the left apical 4-chamber (A) and 2-chamber (B) views, short-axis view (C), and LV chamber cast (D) for the 3-D full volume LV data set of a clinically normal dog. After manual definition of 5 reference points, automated software for detection of the endocardial border creates a cast of the LV and calculates LV volume. Accuracy of detection of the endocardial border was assessed and adjusted manually in each frame. See Figure I for key.
Citation: American Journal of Veterinary Research 77, 11; 10.2460/ajvr.77.11.1211
Contrast-enhanced MDCT
Contrast-enhanced MDCT images were acquired with a 64-row detector CT scanner.f Dogs were placed in sternal recumbency with ECG electrodes attached to the forelimbs and left hind limb. A retrospective ECG-gated MDCT scan was performed with a 3-phase injection of contrast mediumg administered by use of a dual-barrel power injector. The 3-phase injection included an initial fast flow rate of contrast medium (5 mL/s), a second slower rate of contrast medium (2 mL/s), and a final fast rate of sterile saline (0.9% NaCl) solution (5 mL/s), with a total volume of contrast agent of 1 mL/kg and total volume of saline solution of 0.5 mL/kg. Automated bolus-tracking initiated the scanning when 180 Hounsfield units were detected in the ascending aorta. The MDCT scans were performed with the following scanning variables: collimation, 0.5 mm; reconstruction interval, 0.5 mm; gantry rotation speed, 350 milliseconds; 120 kV; 200 to 500 mA; pitch factor, 0.829; and tilt, 0°.
Images were transferred to a workstation and analyzed with LV functional analysis software.h Images were reconstructed in 10% increments of the R–R interval. Multiplanar reconstruction images were created to display standard long- and short-axis views of the LV, rather than transverse, sagittal, and dorsal reconstructions of the thoracic cavity. The functional analysis software displayed the 4-chamber, 2-chamber, and short-axis views of the LV with automated identification of regions of interest for contrast detection (Figure 3). The region of interest for short-axis slices was inspected and adjusted to ensure detection of the entire LV cavity. After adjustments were performed, the software computed the volume by use of the Simpson MOD at each 10% interval and displayed EDV, ESV, SV, and EF.
Reconstructed, ECG-gated, contrast-enhanced cardiac MDCT images of multiplanar reformatting for a 4-chamber slice (A), orthogonal 2-chamber slice (B), cast of the LV chamber (C), and short-axis view (D) of a clinically normal dog. After manual definition of the region of interest, automated software for contrast detection colors the LV cavity (red), creates a cast of the LV, and calculates LV volume. Accuracy of detection of the endocardial border was assessed and adjusted manually at each 10% increment of the R–R interval.
Citation: American Journal of Veterinary Research 77, 11; 10.2460/ajvr.77.11.1211
Statistical analysis
Statistical analysis was performed with commercial software.i Data were assessed for normality by use of the Kolmogorov-Smirnov normality test; data were reported as mean ± SD when normally distributed and as median and range when not normally distributed. Indexed volumes were normally distributed, whereas heart rates, raw volumes, and EF were not. Heart rates during acquisition of echocardiographic and MDCT images were compared by use of a Wilcoxon matched-pairs signed rank test. A 1-way ANOVA with Tukey multiple comparisons test was used to assess differences in values (which were indexed to BSA and body weight) for EDV, ESV, SV, and EF obtained via echocardiography and MDCT. The relationship between volumes and EF obtained with echocardiography and MDCT was analyzed by use of linear regression analysis and the Spearman rank correlation coefficient. Bland-Altman analysis was performed to calculate limits of agreement and systematic errors between echocardiographic methods and MDCT. Significance was set at values of P < 0.05.
Results
Echocardiographic and MDCT examinations of excellent quality were obtained for all dogs. Median heart rate during acquisition of echocardiographic images (75.5 beats/min; range, 65 to 121 beats/min) was not significantly (P = 0.44) different from the median heart rate during acquisition of MDCT images (79 beats/min; range, 67 to 104 beats/min). Raw LV volumes (EDV, ESV, and SV) and EF were calculated (Table 1). Scatter plots of LV volumes for each echocardiographic method had moderate to high correlation with volumes for MDCT, depending on the echocardiographic method (Figure 4). The LV volumes obtained by use of RTBPE, RT3DE, and M-mode echocardiography had the strongest correlations with MDCT for LV volumes (r = 0.97, 0.95, and 0.95, respectively).
Linear regression analysis of end-diastolic (black circles) and end-systolic (white circles) LV volumes measured by use of RTBPE (A), RT3DE (B), M-mode and the Teichholz equation (C), 2-D biplane MOD (D), 2-D single-plane MOD (E), and 2-D single-plane area-length (A-L [F]) methods, compared with LV volumes measured by use of MDCT.
Citation: American Journal of Veterinary Research 77, 11; 10.2460/ajvr.77.11.1211
Median (range) values for LV volumes and EF measured by use of MDCT and various echocardiographic methods.
| 2-D | 3-D | |||||||
|---|---|---|---|---|---|---|---|---|
| Variable | MDCT | l-D M-mode | Single-plane MOD | Single-plane A-L | Biplane MOD | Biplane A-L | Real-time biplane | Real-time 3-D |
| EDV (mL) | 33. S | 35.3 | 34.9 | 36.9 | 37.3 | 38.3 | 31.3 | 32.3 |
| (30.0–72.0) | (25.4–71.4) | (25.4–53.9) | (26.2–57.5) | (25.7–65.7) | (33.2–61.7) | (23.9–59.8) | (23.2–53.1) | |
| ESV (mL) | 19.0 | 19.5 | 21.1 | 22.1 | 19.9 | 24.0 | 19.6 | 19.5 |
| (17.0–36.0) | (14.4–34.1) | (17.4–39.5) | (18.0–42.0) | (16.8–36.7) | (16.9–34.4) | (15.5–33.2) | (13.7–31.5) | |
| SV (mL) | 14.5 | 16.8 | 13.7 | 14.8 | 18.1 | 15.7 | 11.2 | 12.9 |
| (13.0–36.0) | (7.9–37.3) | (8.0–27.4) | (8.2–30.5) | (8.9–37.4) | (11.2–27.3) | (8.4–26.7) | (9.4–23.4) | |
| EF (%) | 44.5 | 48.1 | 38.7 | 39.1 | 46.1 | 41.3 | 41.7 | 40.9 |
| (42.0–50.0) | (31.2–52.3) | (26.0–52.0) | (26.1–54.0) | (34.7–58.3) | (31.9–49.0) | (35.3–46.8) | (38.4–45.0) | |
A-L = Area-length.
The EDV, ESV, and SV results indexed to body weight and BSA were summarized (Table 2). Regardless of indexing method, no significant differences were found between any methods of measurement for indexed values of EDV, ESV, or SV.
Mean ± SD values for LV volume indexed to body weight (IBW) and BSA (IBSA) measured by use of MDCT and various echocardiographic methods.
| 2-D | 3-D | |||||||
|---|---|---|---|---|---|---|---|---|
| Variable | MDCT | l-D M-mode | Single-plane MOD | Single-plane A-L | Biplane MOD | Biplane A-L | Real-time biplane | Real-time 3-D |
| EDV-IBW (mL/kg) | 2.75 ± 0.36 | 2.69 ± 0.73 | 2.63 ± 0.60 | 2.77 ± 0.62 | 2.90 ± 0.56 | 2.93 ± 0.63 | 2.52 ± 0.41 | 2.39 ± 0.49 |
| ESV-IBW (mL/kg) | 1.53 ± 0.24 | 1.46 ± 0.37 | 1.63 ± 0.40 | 1.70 ± 0.43 | 1.56 ± 0.35 | 1.74 ± 0.46 | 1.48 ± 0.30 | 1.40 ± 0.30 |
| SV-IBW (mL/kg) | 1.24 ± 0.18 | 1.23 ± 0.43 | 1.00 ± 0.30 | 1.10 ± 0.32 | 1.34 ± 0.32 | 1.21 ± 0.28 | 1.04 ± 0.20 | 1.00 ± 0.21 |
| ESV-IBSA (mL/m2) | 66.5 ± 8.10 | 64.4 ± 14.60 | 62.7 ± 8.14 | 66.1 ± 8.60 | 69.8 ± 9.80 | 69.7 ± 7.26 | 60.9 ± 7.78 | 58.5 ± 8.26 |
| ESV-IBSA (mL/m2) | 36.7 ± 2.53 | 34.8 ± 5.85 | 38.8 ± 6.80 | 40.6 ± 7.50 | 37.2 ± 5.39 | 41.4 ± 6.66 | 35.6 ± 4.19 | 34.1 ± 4.83 |
| SV-IBSA (mL/m2) | 30.2 ± 5.82 | 29.7 ± 10.10 | 23.9 ± 6.26 | 25.5 ± 6.96 | 32.6 ± 7.96 | 28.4 ± 4.62 | 25.3 ± 5.48 | 24.4 ± 4.19 |
A-L = Area – length.
Results for correlation and bias analysis when EDV and ESV were assessed separately were determined (Table 3). Values obtained for EDV by use of RTBPE, 2-D biplane MOD, and RT3DE and for ESV by use of RTBPE, RT3DE, and M-mode echocardiography had the highest correlations with results for MDCT. The weakest correlations with MDCT were obtained with biplane area-length for EDV and single-plane area-length for ESV. For EDV and ESV, measurements obtained from biplane methods had higher correlations with results for MDCT than did those for single-plane methods. Additionally, measurements obtained by use of MOD had higher correlations with MDCT than did those obtained by use of the area-length method. Values for EF for any of the echocardiographic methods did not correlate highly with results for MDCT.
Results of correlation and Bland-Altman analyses for EDV and ESV between various echocardiographic methods and MDCT.
| EDV | ESV | |||||||
|---|---|---|---|---|---|---|---|---|
| Correlation | Bland-Altman | Correlation | Bland-Altman | |||||
| r | P value | Bias | SD | r | P value | Bias | SD | |
| l-D | ||||||||
| M-mode | 0.92 | 0.01 | 1.72 | 6.93 | 0.92 | 0.01 | 1.43 | 3.14 |
| 2-D | ||||||||
| Single-plane MOD | 0.91 | 0.01 | 3.60 | 8.61 | 0.65 | 0.16 | –0.90 | 6.61 |
| Single-plane A-L | 0.91 | 0.01 | 1.31 | 8.02 | 0.63 | 0.18 | –2.04 | 7.12 |
| Biplane MOD | 0.93 | 0.01 | –1.70 | 6.64 | 0.77 | 0.07 | –0.02 | 5.16 |
| Biplane A-L | 0.85 | 0.03 | −0.48 | 10.00 | 0.77 | 0.07 | –2.22 | 5.01 |
| 3-D | ||||||||
| Real-time biplane MOD | 0.96 | 0.01 | 3.73 | 5.19 | 0.96 | 0.01 | 0.74 | 2.14 |
| Real-time 3-D | 0.92 | 0.01 | 6.25 | 7.71 | 0.96 | 0.01 | 1.98 | 2.57 |
Values were considered significant at P < 0.05.
A-L = Area – length.
Results of Bland-Altman analyses were plotted (Figure 5). In comparison to MDCT, RT3DE slightly underestimated EDV, ESV, and EF, with a bias of 6 mL, 2 mL, and 3.4%, respectively. Underestimation was also found with RTBPE, compared with MDCT, although to a lesser degree than for RT3DE (bias of 4 mL, 1 mL, and 3.7% for EDV, ESV, and EF, respectively).
Graphs of Bland-Altman analysis of the differences between end-diastolic (black circles) and end-systolic (white circles) LV volumes measured by use of RTBPE (A), RT3DE (B), M-mode and the Teichholz equation (C), 2-D biplane MOD (D), 2-D single-plane MOD (E), and 2-D single-plane area-length (F) methods, compared with LV volumes measured by use of MDCT. Mean bias is indicated by the solid line, and limits of agreement (bias ± [1.96•SD]) are indicated by the dotted lines.
Citation: American Journal of Veterinary Research 77, 11; 10.2460/ajvr.77.11.1211
Discussion
In the study reported here, measurements of LV volume and function obtained by use of 3-D echocardiography were highly correlated with measurements obtained by use of MDCT in clinically normal dogs. Additionally, results of the study indicated that 3-D imaging modes, including RT3DE and RTBPE, performed slightly better than did 2-D imaging modes when compared with MDCT. Bland-Altman analysis reflected that 3-D echocardiographic methods minimally underestimated EDV, ESV, and EF in comparison to results for MDCT, although with smaller limits of agreement than for 2-D echocardiographic methods. The M-mode and 2-D echocardiography single-plane MOD measurements also underestimated LV volumes, whereas 2-D echocardiography biplane MOD and single-plane area-length slightly overestimated volumes, compared with results for MDCT. For all methods, the bias was small (range, −0.4 to 3.5 mL), and the narrowest 95% limits of agreement were found with RTBPE for LV volumes. Interestingly, values for EF for any of the echocardiographic methods did not correlate highly with results of MDCT, and this may have been attributable to the small sample size of the study and the modest range of EF values in this population of clinically normal dogs. Small differences among methods for values of EDV and ESV may cause larger differences when applied to the EF formula and thus may yield a less robust correlation than the individual variables alone.
Raw values for LV volumes as well as those indexed to both body weight and BSA were reported. It is difficult to compare RT3DE and MDCT LV volumes reported in the present study with those obtained in previous veterinary studies because most investigators reported only raw volumes.9,11,19,20 Most of those studies involved the use of a small number of clinically normal dogs of approximately the same size; therefore, it was likely deemed unnecessary to report indexed volumes. In the present study, RT3DE values for EDV and ESV indexed to BSA (mean, 58.5 and 34.1, respectively) were similar to RT3DE values for EDV and ESV indexed to BSA for female humans (mean, 58 and 23, respectively21), taking into account that the study reported here was conducted in anesthetized dogs receiving an esmolol infusion. There is still debate among veterinary and human cardiologists as to the best method for indexing LV dimensions and volumes. Some argue that indexing volume (which is a 3-D variable) to BSA (which is a 2-D variable) violates the geometric relationship between them.22 Allometric scaling models use the equation Y= aMb, where Y represents the measure of heart size, a is a constant, M is body weight, and b is the exponential scaling constant. Generally, when Y is a volume measure, the exponent for body weight is 1, rather than one-third for linear measures or two-thirds for area measures.23 This relationship perhaps remains unclear because some investigators assessing allometric relationships of LV volumes have reported exponents other than 1 following logarithmic transformation to obtain allometric equations.24 Future studies with large populations of dogs are warranted to better delineate the allometric relations of cardiac volumes and body weight and to ultimately help differentiate clinically normal animals from abnormal patients. The authors believed this population of dogs was too small to attempt to assess these relationships, and therefore, we elected to report the raw volumes and results for both common indexing methods to allow other investigators the opportunity to compare values.
The fact that the RTBPE 3DE method had the highest correlation with MDCT for EDV, ESV, and EF was an interesting finding of the present study. Real-time biplane echocardiography uses 3-D pyramidal data to simultaneously display the 4- and 2-chamber left apical views from the same cardiac cycle. The quantification software for RTBPE is faster and simpler than that of RT3DE for obtaining LV volumes and EF because of manual adjustments required with the automated 3-D echocardiographic quantification software. Although the equivalent performance of RTBPE and RT3DE was an unexpected finding in the study reported here, investigators of another recent study11 reported stronger LV volume correlations for real-time triplane echocardiography than for RT3DE when compared with results for cardiac MRI. This was an interesting finding in that study11 and the study reported here because although RTBPE involves the use of 3-D data, it ultimately uses geometric shape assumptions to calculate LV volumes, whereas RT3DE as a volumetric method does not use shape assumptions. The lack of superior performance by RT3DE may have been attributable to poor definition of the endocardial border, user error in adjustments to the endocardial border, or inaccuracies in the software algorithm used to calculate volumes. Alternatively, the small sample size may have rendered the study underpowered to identify a superior performance of RT3DE over RTBPE. Manual adjustments made to the LV cast with the RT3DE software were time consuming, and it is clinically important to mention that the measurements made with the simpler, faster RTBPE method were as accurate as those obtained by use of MDCT.
The superiority of 3-D echocardiographic methods, including RTBPE and RT3DE, over 2-D biplane methods was less surprising and likely attributable to the difficulty in obtaining a nonforeshortened, true orthogonal 2-chamber view from the left apical window in dogs. Real-time biplane echocardiography overcomes this disadvantage by use of 3-D data acquired to simultaneously display the 4- and 2-chamber views, which allow users to confirm true orthogonal alignment and inclusion of the entire LV apex. Overall, LV volumes obtained with 2-D echocardiography correlated well with results for MDCT, and values for biplane 2-D methods correlated better with results for MDCT than did the corresponding single-plane methods. This observation has been found consistently in studies of humans and relates to the fact that fewer shape assumptions are required when additional views of the ventricle are obtained. Interestingly, the single-plane and biplane MOD methods, compared with the single-plane and biplane area-length methods, had higher correlations with MDCT. The poor correlation of the single-plane and biplane area-length methods with MDCT in this study could have been attributable to the small sample size or to the fact that the MDCT volume calculations were obtained by use of the Simpson method. Therefore, it may be expected that 2-D measurements made by use of the modified Simpson rule would correlate better with MDCT values made by use of the Simpson method.
Estimates of LV volumes obtained by use of the Teichholz method agreed surprisingly well with values obtained by the use of MDCT, which offers support to the validity of the Teichholz equation when the LV is of normal size and shape. In a recent study25 in which investigators compared 2-D echocardiographic LV volumes calculated via the modified Simpson and Teichholz methods with values for dual-source CT, mean EDV, ESV, and SV calculated by use of the Teichholz method were more similar to dual-source CT volumes than were values calculated by use of the modified Simpson method. It is unlikely that the ellipsoid shape assumptions inherent in the Teichholz equation would yield accurate results when the LV cavity changes shape, as has been confirmed for volume overload10 and various cardiomyopathies.26,27 Differences in LV volume estimates obtained by use of 2-D and 3-D echocardiography typically are small for clinically normal hearts, and it is suspected that the superiority of 3-D echocardiography would become more apparent with altered ventricular size or function. This has been reported for hypertrophic cardiomyopathy,28 dilated cardiomyopathy,29 and LV aneurysms30 in humans and sheep. This most likely occurs as a result of further deviation from the geometric shape assumptions inherent in 2-D echocardiographic volume calculations. Thus, when the LV has a relatively normal shape, the 2-D echocardiographic MOD performs well for estimation of a 3-D volume from single-plane or biplane images, but performance and accuracy of 2-D echocardiography decline as the LV remodels during disease. Because of the lack of obligate shape assumptions, 3-D echocardiographic methods overcome this limitation of 2-D echocardiographic methods.
The present study had limitations. First, use of anesthesia and an esmolol infusion likely altered cardiac function, although randomization of imaging order should have protected against an affect of anesthesia time on comparisons. Second, the population of dogs was small, and although there was some variation in body weight, there was homogeneity in age, sex, and overall health. Additional studies involving the use of RT3DE and RTBPE in a large population of dogs will be needed to assess volumes in nonanesthetized patients and explore allometric relationships to establish reference values for these methods.
There has been limited evaluation of RT3DE in the field of veterinary medicine,9–13 although it is anticipated that the use of this modality will increase as its costs decline and technology continues to advance. The use of 3-D echocardiography (both RT3DE and RTBPE methods) to measure LV volumes and function provided excellent agreement with values obtained by use of MDCT in clinically normal dogs. Additional studies validating the use of 3-D echocardiographic volume measurements in disease states with ventricular remodeling are warranted to further verify the use of this imaging modality in veterinary patients.
Acknowledgments
No third-party funding or support was received in connection with this study or the writing or publication of the manuscript. The authors declare that there were no conflicts of interest.
The authors thank Amy Berry, Robyn Panico, and Jason Wiest for technical assistance.
ABBREVIATIONS
| BSA | Body surface area |
| EDV | End-diastolic volume |
| EF | Ejection fraction |
| ESV | End-systolic volume |
| LV | Left ventricle |
| MDCT | Multidetector row CT |
| MOD | Method of disks |
| RT3DE | Real-time 3-D echocardiography |
| RTBPE | Real-time biplane echocardiography |
| SV | Stroke volume |
Footnotes
Brant R. Inference for Means. University of British Columbia, Department of Statistics. Available at: www.stat.ubc.ca/∼rollin/stats/ssize/n2.html. Accessed Oct 1, 2012.
iE33, Philips Ultrasound, Bothell, Wash.
Xcelera Express, Philips Ultrasound, Bothell, Wash.
QLAB 3DQ, version 9.0, Philips Ultrasound, Bothell, Wash.
QLAB 3DQA, version 9.0, Philips Ultrasound, Bothell, Wash.
Toshiba Aquilion 64 CT, Toshiba America Medical Systems Inc, Tustin, Calif.
Isovue 370, Bracco Diagnostics Inc, Princeton, NJ.
Vitrea workstation, version 6.3.2, Vital Images Inc, Minnetonka, Minn.
GraphPad Prism, version 6.04, GraphPad Software Inc, La Jolla, Calif.
References
1. Atkins C, Bonagura J, Ettinger S, et al. Guidelines for the diagnosis and treatment of canine chronic valvular heart disease. J Vet Intern Med 2009; 23:1142–1150.
2. Dukes-McEwan J, Borgarelli M, Tidholm A, et al. Proposed guidelines for the diagnosis of canine idiopathic dilated cardiomyopathy. J Vet Cardiol 2003; 5:7–19.
3. Teichholz LE, Kreulen T, Herman MV, et al. Problems in echocardiographic volume determinations: echocardiographic-angiographic correlations in the presence of absence of asynergy. Am J Cardiol 1976; 37:7–11.
4. Lang RM, Badano LP, Mor-Avi V, et al. Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr 2015; 28:1–39.
5. Chukwu EO, Barasch E, Mihalatos DG, et al. Relative importance of errors in left ventricular quantitation by two-dimensional echocardiography: insights from three-dimensional echocardiography and cardiac magnetic resonance imaging. J Am Soc Echocardiogr 2008; 21:990–997.
6. Dorosz JL, Lezotte DC, Weitzenkamp DA, et al. Performance of 3-dimensional echocardiography in measuring left ventricular volumes and ejection fraction: a systematic review and meta-analysis. J Am Coll Cardiol 2012; 59:1799–1808.
7. Sapin PM, Schroder KM, Gopal AS, et al. Comparison of two- and three-dimensional echocardiography with cineventriculography for measurement of left ventricular volume in patients. J Am Coll Cardiol 1994; 24:1054–1063.
8. Jenkins C, Bricknell K, Hanekom L, et al. Reproducibility and accuracy of echocardiographic measurements of left ventricular parameters using real-time three-dimensional echocardiography. J Am Coll Cardiol 2004; 44:878–886.
9. Tidholm A, Westling AB, Hoglund K, et al. Comparisons of 3-, 2-dimensional, and M-mode echocardiographical methods for estimation of left chamber volumes in dogs with and without acquired heart disease. J Vet Intern Med 2010; 24:1414–1420.
10. Ljungvall I, Hoglund K, Carnabuci C, et al. Assessment of global and regional left ventricular volume and shape by real-time 3-dimensional echocardiography in dogs with myxomatous mitral valve disease. J Vet Intern Med 2011; 25:1036–1043.
11. Meyer J, Wefstaedt P, Dziallas P, et al. Assessment of left ventricular volumes by use of one-, two-, and three-dimensional echocardiography versus magnetic resonance imaging in healthy dogs. Am J Vet Res 2013; 74:1223–1230.
12. Tidholm A, Bodegard-Westling A, Hoglund K, et al. Comparisons of 2- and 3-dimensional echocardiographic methods for estimation of left atrial size in dogs with and without myxomatous mitral valve disease. J Vet Intern Med 2011; 25:1320–1327.
13. Tidholm A, Hoglund K, Haggstrom J, et al. Left atrial ejection fraction assessed by real-time 3-dimensional echocardiography in normal dogs and dogs with myxomatous mitral valve disease. J Vet Intern Med 2013; 27:884–889.
14. Sapin PM, Schröeder KM, Gopal AS, et al. Three-dimensional echocardiography: limitations of apical biplane imaging for measurement of left ventricular volume. J Am Soc Echocardiogr 1995; 8:576–584.
15. Greupner J, Zimmermann E, Grohmann A, et al. Head-to-head comparison of left ventricular function assessment with 64-row computed tomography, biplane left cineventriculography, and both 2- and 3-dimensional transthoracic echocardiography: comparison with magnetic resonance imaging as the reference standard. J Am Coll Cardiol 2012; 59:1897–1907.
16. Mahnken AH, Muhlenbruch G, Koos R, et al. Automated vs. manual assessment of left ventricular function in cardiac multidetector row computed tomography: comparison with magnetic resonance imaging. Eur Radiol 2006; 16:1416–1423.
17. Raman SV, Shah M, McCarthy B, et al. Multi-detector row cardiac computed tomography accurately quantifies right and left ventricular size and function compared with cardiac magnetic resonance. Am Heart J 2006; 151:736–744.
18. Guo YK, Yang ZG, Ning G, et al. Sixty-four-slice multidetector computed tomography for preoperative evaluation of left ventricular function and mass in patients with mitral regurgitation: comparison with magnetic resonance imaging and echocardiography. Eur Radiol 2009; 19:2107–2116.
19. Sieslack AK, Dziallas P, Nolte I, et al. Comparative assessment of left ventricular function variables determined via cardiac computed tomography and cardiac magnetic resonance imaging in dogs. Am J Vet Res 2013; 74:990–998.
20. Henjes CR, Hungerbuhler S, Bojarski IB, et al. Comparison of multi-detector row computed tomography with echocardiography for assessment of left ventricular function in healthy dogs. Am J Vet Res 2012; 73:393–403.
21. Aune E, Baekkevar M, Rodevand O, et al. Reference values for left ventricular volumes with real-time 3-dimensional echocardiography. Scand Cardiovasc J 2010; 44:24–30.
22. Dewey FE, Rosenthal D, Murphy DJ Jr, et al. Does size matter? Clinical applications of scaling cardiac size and function for body size. Circulation 2008; 117:2279–2287.
23. Cornell CC, Kittleson MD, Della Torre P, et al. Allometric scaling of M-mode cardiac measurements in normal adult dogs. J Vet Intern Med 2004; 18:311–321.
24. Zong P, Zhang L, Shaban NM, et al. Left heart chamber quantification in obese patients: how does larger body size affect echocardiographic measurements? J Am Soc Echocardiogr 2014; 27:1267–1274.
25. Lee M, Park N, Lee S, et al. Comparison of echocardiography with dual-source computed tomography for assessment of left ventricular volume in healthy Beagles. Am J Vet Res 2013; 74:62–69.
26. Vandenbossche JL, Kramer BL, Massie BM, et al. Two-dimensional echocardiographic evaluation of the size, function and shape of the left ventricle in chronic aortic regurgitation: comparison with radionuclide angiography. J Am Coll Cardiol 1984; 4:1195–1206.
27. Wandt B, Bojo L, Tolagen K, et al. Echocardiographic assessment of ejection fraction in left ventricular hypertrophy. Heart 1999; 82:192–198.
28. Bicudo LS, Tsutsui JM, Shiozaki A, et al. Value of real time three-dimensional echocardiography in patients with hypertrophic cardiomyopathy: comparison with two-dimensional echocardiography and magnetic resonance imaging. Echocardiography 2008; 25:717–726.
29. Gutierrez-Chico JL, Zamorano JL, Perez de Isla L, et al. Comparison of left ventricular volumes and ejection fractions measured by three-dimensional echocardiography versus by two-dimensional echocardiography and cardiac magnetic resonance in patients with various cardiomyopathies. Am J Cardiol 2005; 95:809–813.
30. Qin JX, Jones M, Shiota T, et al. Validation of real-time three-dimensional echocardiography for quantifying left ventricular volumes in the presence of a left ventricular aneurysm: in vitro and in vivo studies. J Am Coll Cardiol 2000; 36:900–907.
