Ml For End-systolic Volume Research Articles

Assessment of deep learning segmentation for real-time free-breathing cardiac magnetic resonance imaging at rest and under exercise stress

In recent years, a variety of deep learning networks for cardiac MRI (CMR) segmentation have been developed and analyzed. However, nearly all of them are focused on cine CMR under breathold. In this work, accuracy of deep learning methods is assessed for volumetric analysis (via segmentation) of the left ventricle in real-time free-breathing CMR at rest and under exercise stress. Data from healthy volunteers (n = 15) for cine and real-time free-breathing CMR at rest and under exercise stress were analyzed retrospectively. Exercise stress was performed using an ergometer in the supine position. Segmentations of two deep learning methods, a commercially available technique (comDL) and an openly available network (nnU-Net), were compared to a reference model created via the manual correction of segmentations obtained with comDL. Segmentations of left ventricular endocardium (LV), left ventricular myocardium (MYO), and right ventricle (RV) are compared for both end-systolic and end-diastolic phases and analyzed with Dice’s coefficient. The volumetric analysis includes the cardiac function parameters LV end-diastolic volume (EDV), LV end-systolic volume (ESV), and LV ejection fraction (EF), evaluated with respect to both absolute and relative differences. For cine CMR, nnU-Net and comDL achieve Dice’s coefficients above 0.95 for LV and 0.9 for MYO, and RV. For real-time CMR, the accuracy of nnU-Net exceeds that of comDL overall. For real-time CMR at rest, nnU-Net achieves Dice’s coefficients of 0.94 for LV, 0.89 for MYO, and 0.90 for RV and the mean absolute differences between nnU-Net and the reference are 2.9 mL for EDV, 3.5 mL for ESV, and 2.6% for EF. For real-time CMR under exercise stress, nnU-Net achieves Dice’s coefficients of 0.92 for LV, 0.85 for MYO, and 0.83 for RV and the mean absolute differences between nnU-Net and reference are 11.4 mL for EDV, 2.9 mL for ESV, and 3.6% for EF. Deep learning methods designed or trained for cine CMR segmentation can perform well on real-time CMR. For real-time free-breathing CMR at rest, the performance of deep learning methods is comparable to inter-observer variability in cine CMR and is usable for fully automatic segmentation. For real-time CMR under exercise stress, the performance of nnU-Net could promise a higher degree of automation in the future.

Three-Tesla magnetic resonance imaging of left ventricular volume and function in comparison with computed tomography and echocardiography.

This study investigated the correlation between 3-Tesla magnetic resonance imaging (MRI) and 256 multiple-slice computed tomography (MSCT) or 2-dimensional echocardiography (ECHO) in evaluating left ventricle. Forty patients were retrospectively enrolled to undergo cardiac MSCT, 3-Tesla MRI and 2-dimensional ECHO within 1 week. The end-diastolic (EDV) and end-systolic volume (ESV), stroke volume (SV) and ejection fraction (EF) were analyzed and compared. MSCT was highly significantly correlated with MRI. Compared with MRI, MSCT slightly overestimated ESV for about 8.7 mL, but slightly underestimated EF and SV for about 6.8% and 5.8 mL, respectively. A high consistency existed between MSCT and MRI, with the 95% limit of agreement (-19.6, 25.4) mL for EDV, (-2.6,20.1) mL for ESV, (-28.3,16.6) mL for SV, and (-18.8%,5.1) % for EF. ECHO was also significantly correlated with MRI. The ECHO slightly underestimated the left ventricular function compared with MRI, with an underestimation of 9.4 mL for EDV, 3.5 mL for ESV, 5.8 mL for SV and 1.0% for EF. A wider agreement limit existed between MRI and ECHO. MSCT has a better correlation and agreement relationship with MRI parameters than 2-dimensional ECHO in assessing the left ventricle and may serve as a possible alternative to MRI.

A Pilot Project to Identify Simple Echocardiographic Tools as an Alternative to Cardiac Magnetic Resonance Imaging to Predict a Reduced Right Ventricular Ejection Fraction in Patients with Repaired Tetralogy of Fallot

Background: Right ventricular (RV) dysfunction is an important predictor of adverse events after tetralogy of Fallot (TOF) repair. A study comparing echocardiography with cardiac magnetic resonance (CMR) might provide simple tools for their serial inexpensive assessment. Methods: Echocardiographic RV functional parameters including fractional area change (FAC), rate of pressure rise (dP/dt), free wall strain, myocardial performance index (MPI), tricuspid annular plane systolic excursion (TAPSE), and tissue Doppler S' velocity in patients with repaired TOF were correlated with CMR-derived ejection fraction (EF) using receiver operator characteristic (ROC) curves. Bland–Altman plots analyzed agreement between the RV end-diastolic volumes (EDVs), end-systolic volumes (ESVs), and EF derived from three-dimensional echocardiography (3DE) and CMR. Results: Eighteen patients were included. The echocardiographic parameters that showed a good positive correlation with CMR-derived EF were FAC (r = 0.851, P = 0.001), dP/dt (r = 0.730, P = 0.001), and free wall strain (r = −0.660, P = 0.003). ROC curve analysis provided a cutoff value for FAC (<36.6), dP/dt (<370 mmHg/s), free wall strain (<−18.5) and MPI (>0.565) to predict EF <45%. TAPSE and S' velocity had poor correlation. Correlation was strong between 3DE and CMR-derived EDV (r = 0.938, P <0.001), ESV (r = 0.936, P <0.001) and EF (r = 0.916, P <0.001). 3DE underestimated volumes compared to CMR with a mean bias of − 31.78 ± 18.8 ml for EDV and −17.28 ± 11.6 ml for ESV, but EF was not affected (mean bias − 1% ± 3.7%). Conclusions: RV free wall strain, FAC and dP/dt were simple tools with good accuracy to predict RV EF <45% in patients with repaired TOF. TAPSE and S' velocity assessed basal longitudinal function without considering the dysfunctional outflow tract and showed poor correlation. EF assessment by 3DE was a good alternative to CMR. Underestimation of volumes by 3DE might restrict its use in timing surgery.

Effect of age and sex on fully automated deep learning assessment of left ventricular function, volumes, and contours in cardiac magnetic resonance imaging.

Deep learning algorithms for left ventricle (LV) segmentation are prone to bias towards the training dataset. This study assesses sex- and age-dependent performance differences when using deep learning for automatic LV segmentation. Retrospective analysis of 100 healthy subjects undergoing cardiac MRI from 2012 to 2018, with 10 men and women in the following age groups: 18-30, 31-40, 41-50, 51-60, and 61-80years old. Subjects underwent 1.5T, 2D CINE SSFP MRI. 35 pathologic cases from local clinical exams and the SCMR 2015 consensus contours dataset were also analyzed. A fully convolutional network (FCN) similar to U-Net trained on the U.K. Biobank was used to automatically segment LV endocardial and epicardial contours. FCN and manual segmentation were compared using Dice metrics and measurements of end-diastolic volume (EDV), end-systolic volume (ESV), mass (LVM), and ejection fraction (LVEF). Paired t-tests and linear regressions were used to analyze measurement differences with respect to sex and age. Dice metrics (median ± IQR) for n = 135 cases were 0.94 ± 0.04/0.87 ± 0.10 (ED endocardium/ES endocardium). Measurement biases (mean ± SD) among the healthy cohort were -0.3 ± 10.1mL for EDV, -6.7 ± 9.6mL for ESV, 4.6 ± 6.4% for LVEF, and -2.2 ± 11.0g for LVM; biases were independent of sex and age. Biases among the 35 pathologic cases were 0.1 ± 19mL for EDV, -4.8 ± 19mL for ESV, 2.0 ± 7.6% for LVEF, and 1.0 ± 20g for LVM. In conclusion, automatic segmentation by the Biobank-trained FCN was independent of age and sex. Improvements in end-systolic basal slice detection are needed to decrease bias and improve precision in ESV and LVEF.

1180 2D and 3D assessment of the left ventricle volume and ejection fraction in a general population

Abstract Introduction In standard practice, LV volumes and EF are estimated by 2D technique. 3D echocardiographic assessment seems more reliable; however, this method has not yet been validated in the general population. Purpose To validate 3D echocardiography in a large population sample and investigate differences between 2D and 3D LVEF and volumes Methods In The Copenhagen City Heart Study, 4466 echocardiograms were available for analysis. The echocardiograms were obtained during four consecutive heartbeats in both 2D and 3D with GE Vivid E9. Offline analysis was performed on EchoPac v. 201. LVEF was calculated by the modified Simpsons Biplane Auto EF for 2D and by the 4LVQ method for 3D. Results The study included 2090 echocardiograms. The mean 2D LVEF was 57.3 ± 6.1% (IQR 54 - 61%) and 51.7 ± 7.9% (IQR 47 - 57%) by 3D. The mean end-diastolic volume (EDV) and end-systolic volume (ESV) by 2D and 3D techniques were: EDV 2D 106.1 ± 29.6 ml vs EDV 3D 128.2 ± 32.3 ml , ESV 2D 45.7 ± 15.6 ml vs. ESV 3D 45.7 ± 20.7 , p < 0.05 among all variables. The average difference of means between 2D and 3D LVEF was 5.6 ± 11.2%, -22.1 ± 56.8 ml for EDV, and -16.9 ± 32.9 ml for ESV. The correlation coefficient for LVEF was 0.42, EDV 0.76 and for ESV 0.70. Conclusion In our study, we found a significant difference in both LVEF and ventricular volumes when comparing 2D echocardiograms with 3D. 3DE had, in general, lower LVEF, higher EDV and ESV compared to 2D. Table 1: Summary of results Table 1 - Summary of results n = 2090 Variable Min Max Mean IQR (25-75) p-value LVEF, 2D (%) 18 76 57.3 ± 6.1 54-61 < 0.05 LVEF, 3d (%) 13 77 51.7 ± 7.9 47-57 < 0.05 EDV, 2D (ml) 13 275 106.1 ± 29.6 85-123.8 < 0.05 EDV, 3D (ml) 50 270 128.2 ± 32.3 106-148 < 0.05 ESV, 2D (ml) 15 150 45.7 ± 15.6 35-54 < 0.05 ESV, 3D (ml) 13 185 45.7 ± 20.7 48-74 < 0.05 LVEF: left ventricle ejection fraction, EDV: end-diastolic volume, ESV: end systolic volume, IQR: Inter-quartile range Abstract 1180 Figure 1: Correlation and BA-plot

Improved workflow for quantifying left ventricular function via cardiorespiratory-resolved analysis of free-breathing MR real-time cines.

To evaluate the feasibility of a proposed cardiorespiratory-resolved analysis in left ventricular (LV) function quantification from real-time cines in a cohort of cardiac patients. Electrocardiograph (ECG)-free free-breathing real-time cine imaging based on the balanced steady-state free precession technique was performed on short-axis slices of 20 cardiac patients at 3T. K-means cluster segmentation was used to delineate the endocardial contours, from which the LV centroid and cavity area were determined. Respiratory and cardiac signals were respectively resolved from the filtered LV centroid displacement and time-varied LV cavity area to identify end-expiratory end-diastolic (ED) and end-systolic (ES) images. The obtained LV cavity areas and derived volumetric function indices, including ED volume (EDV), ES volume (ESV), stroke volume (SV), and ejection fraction (EF), were compared with those measured from manual analysis using two-tailed paired Student's t-tests, linear regression analyses, and Bland-Altman plots. Interobserver variability was calculated. The LV cavity area was strongly correlated between the proposed and conventional manual methods (r > 0.87) for three representative slices at the base, middle ventricle, and apex. The average differences between the two methods were 0.66 ± 3.22 mL for EDV, -0.02 ± 2.68 mL for ESV, 0.67 ± 3.73 mL for SV, and 0.17 ± 2.30% for EF. All paired measures exhibited strong correlations (r > 0.96) without significant differences (P = 0.38-0.98). Acceptable interobserver variability (0.19-3.55%) and strong correlations (r > 0.96) were shown for all measures between the two observers. The proposed method is feasible for efficient measurement of LV function from real-time cines. 2 Technical Efficacy: Stage 1 J. MAGN. RESON. IMAGING 2017;46:905-914.

Efficient method for analyzing MR real-time cines: Toward accurate quantification of left ventricular function.

To develop and assess an efficient method to identify end-expiratory end-diastolic (ED) and end-systolic (ES) images for accurate quantification of left ventricular (LV) function in real-time cine imaging. ECG-free free-breathing real-time cine imaging was performed on short-axis slices of thirteen healthy volunteers at 3 Tesla. K-means cluster segmentation was applied to delineate the endocardial contour, from which the LV centroid and cavity area were determined in each cine image. LV centroid displacement along the superior-inferior direction was filtered to extract respiratory motion in each slice. End-expiratory ED and ES images were then identified and used for LV function quantification. Accuracy was evaluated with that measured from the slice-matched standard ECG-gated breathhold segmented cines using two-tail paired Student's t-tests, linear regression analyses and Bland-Altman plots. Intra- and interobserver variability was calculated for each imaging technique. Qualitatively, end-expiratory ED and ES images identified with the proposed method agreed with those determined by frame-by-frame visual inspection in 97.5% of cases (P > 0.05). Quantitatively, good agreement of LV function indices between the real-time cine and the standard segmented cine was observed with averaged differences of 0.1 ± 0.9 g for myocardium mass, -0.3 ± 1.0 mL for ED volume, 0.2 ± 1.2 mL for ES volume, -0.2 ± 1.3 mL for stroke volume, and -0.3 ± 0.8% for ejection fraction. Paired LV function values exhibited strong correlation (r ≥ 0.96) and no significant difference (P > 0.05). The real-time cine and the standard segmented cine showed similar intra- (1.2-3.3% and 1.1-2.8%, respectively) and interobserver variability (2.6-6.9% and 1.8-4.8%, respectively) with all P-values > 0.05. All the variability was comparable with published results. Compared with the laborious frame-by-frame visual inspection, as conventionally adopted, the proposed method is efficient in analyzing real-time cines for the accurate quantification of LV function without excessively manual interactions.

Single-beat real-time three-dimensional echocardiographic automated contour detection for quantification of left ventricular volumes and systolic function

To assess the feasibility and accuracy in measuring left ventricular (LV) end-diastolic volume (EDV), end-systolic volume (ESV) and ejection fraction (EF) with Siemens single-beat real-time 3D transthoracic echocardiography. The LV volumes and EF were measured in 3D datasets acquired by six imaging modes (time-1-harmonic (T1H), time-1-fundamental, time-2-harmonic, time-2-fundamental, space-1-harmonic (S1H), and space-1-fundamental) in 41 patients using the automated contouring algorithm and compared with manually corrected 3DE QLAB measurements. The main determinates of the temporal and spatial resolutions of 3D datasets acquired were the fundamental and harmonic modes. Consequently, the S1H mode had the lowest volume rate and highest spatial resolution. Compared with the 3DE QLAB analysis, the S1H mode resulted in the best LV volumes and EF estimates in all patients (0 ± 10 % for EF, -7 ± 44 ml for EDV, -7 ± 39 ml for ESV) and in the 10 patients with correct LV contour tracking according to a visual assessment from the multiplanar reconstruction views in all six modes (0 ± 9 % for EF, -3 ± 23 ml for EDV, -2 ± 14 ml for ESV). The T1H mode was the best alternative. Overall 28 patients (68 %) could be analysed automatically and satisfyingly with the S1H and T1H modes: 0 ± 8 % (EF), 0 ± 27 ml (EDV) and -1 ± 16 ml (ESV). The accuracy of the Siemens automated RT-3D algorithm in measuring LV volumes and EF is significantly influenced by the different imaging modes. The S1H mode may be the preferred 3D acquisition mode, supplemented by the T1H mode in enlarged LVs that do not fit in the S1H acquisition sector.

Accurate computer-aided quantification of left ventricular parameters: experience in 1555 cardiac magnetic resonance studies from the Framingham Heart Study.

Quantitative analysis of short-axis functional cardiac magnetic resonance images can be performed using automatic contour detection methods. The resulting myocardial contours must be reviewed and possibly corrected, which can be time-consuming, particularly when performed across all cardiac phases. We quantified the impact of manual contour corrections on both analysis time and quantitative measurements obtained from left ventricular short-axis cine images acquired from 1555 participants of the Framingham Heart Study Offspring cohort using computer-aided contour detection methods. The total analysis time for a single case was 7.6 ± 1.7 min for an average of 221 ± 36 myocardial contours per participant. This included 4.8 ± 1.6 min for manual contour correction of 2% of all automatically detected endocardial contours and 8% of all automatically detected epicardial contours. However, the impact of these corrections on global left ventricular parameters was limited, introducing differences of 0.4 ± 4.1 mL for end-diastolic volume, -0.3 ± 2.9 mL for end-systolic volume, 0.7 ± 3.1 mL for stroke volume, and 0.3 ± 1.8% for ejection fraction. We conclude that left ventricular functional parameters can be obtained under 5 min from short-axis functional cardiac magnetic resonance images using automatic contour detection methods. Manual correction more than doubles analysis time, with minimal impact on left ventricular volumes and ejection fraction.

Robustness of a new three-dimensional echocardiographic algorithm for left ventricular volume and ejection fraction quantification: experts vs. novices

We evaluated the ability of a new simplified algorithm for three-dimensional echocardiography (3DE) left ventricular (LV) measurements with minimal operator interaction to be reproducible and robust, independently of the experience. A total of 163 subjects were investigated using two-dimensional echocardiography (2DE) and 3DE. The 3D data sets were blindly analysed offline by novice investigators and experts. A subgroup of 30 patients was assessed using cardiac magnetic resonance imaging (CMRI) to compare end-diastolic volume (EDV), end-systolic volume (ESV), and ejection fraction (EF) obtained by 2DE, 3DE, and CMRI. Intra-observer and inter-observer variabilities of 2DE and 3DE measurements were evaluated according to level of experience. Mean time analysis of 3DE data was 23.2 ± 6.3s for the novice and 26.1 ± 4.1 s for the expert (P = ns). Correlations (r) and mean error measurements (MEM) between 3DE analysis by experts and novices were 0.91 and -3.5 mL for EDV, 0.97 and 4.3 mL for ESV, and 0.91 and -2.6% for EF, respectively. Correlations between 3DE and CMRI were good with low variability and greater agreement when compared with those between 2DE and CMRI. For the novice, MEM was -21.3 mL for EDV, -15.0 mL for ESV, and 2.3% for EF. MEM and 95% confidence intervals were wider for 2DE vs. CMRI than for 3DE vs. CMRI in relation to both expert and novice. This new semi-automated algorithm of LV endocardial border detection based on 3DE data appears suitable for clinical use by either expert or novice investigators with greater reproducibility and time of analysis than 2DE.

Test-Retest Variability of Volumetric Right Ventricular Measurements Using Real-Time Three-Dimensional Echocardiography

Substantial variability in sequential echocardiographic right ventricular (RV) quantification may exist. Interobserver and intraobserver values are well known, but acquisition (test-retest) variability has been rarely assessed. The objective of this study was to determine the test-retest variability of sequential RV volume and ejection fraction (EF) measurements by real-time three-dimensional echocardiography in patients with congenital heart disease and healthy controls. Twenty-eight participants (21 patients with congenital heart disease, seven healthy controls; mean age, 30 ± 14 years; 43% men) underwent a series of three echocardiographic studies. To obtain interobserver and intraobserver test-retest variability, two sonographers acquired sequential RV data sets in each participant during one outpatient visit. RV volumetric quantification was done using semiautomated three-dimensional border detection. The variability data were analyzed using correlation coefficients, Bland-Altman analysis, and coefficients of variation. Absolute mean differences for sequential intraobserver acquisitions were 12 ± 12 mL for end-diastolic volume, 7 ± 6 mL for end-systolic volume, and 4 ± 3% for EF. Interobserver and intraobserver test-retest variability, respectively, were 7% and 7% for RV end-diastolic volume, 14% and 7% for end-systolic volume, and 8% and 6% for EF. Good test-retest variability, besides the practical nature of real-time three-dimensional echocardiography for RV volume and EF assessment, makes it a valuable technique for serial follow-up. Although it may be challenging to diminish all factors that can influence echocardiographic examination for serial follow-up, standardization of RV size and functional measurements should be a goal to produce more interchangeable data.

Comparative values of gated blood-pool SPECT and CMR for ejection fraction and volume estimation

Gated blood-pool single-photon emission computed tomography (GBPS) was compared with cardiac magnetic resonance (CMR) for the measurement of left ventricular (LV) and right ventricular (RV) ejection fractions (EF) and volumes [end-diastolic volume (EDV) or end-systolic volume (ESV)] in a mixed population. Thirty patients (70% men; mean age: 61±14 years) referred for various symptoms or heart diseases, predominantly ischemic, were included. GBPS data were analyzed using segmentation software described earlier based on the watershed algorithm. CMR images were acquired for both ventricles at the same time using a steady-state-free precession sequence and short-axis views. No compensation for papillary muscles was used. LVEF and RVEF and volumes were assessed with GBPS and CMR and were compared. LVEF and volumes were correlated (P<0.001). The difference in LVEF between GBPS and CMR was not significant (P=0.063). The limits of agreement were close for LVEF (-11 to 15%) and wider for LV volumes (-82 to 11 ml for EDV and -52 to 15 ml for ESV), with higher volume values obtained with CMR (mean differences of 36±24 ml for EDV and 19±17 ml for ESV). The RVEF and volumes assessed by GBPS and CMR were correlated (P<0.001). The difference in RVESV between GBPS or CMR was not significant (P=0.136). The limits of agreement were relatively close for all RV parameters (-15 to 8% for EF; -44 to 22 ml for EDV, and -25 to 21 ml for ESV). In 24 patients without valvulopathy or shunt, the difference between LV stroke volume and RV stroke volume was lower with GBPS than with CMR (9±14 ml and 18±13 ml, respectively, with P=0.027). GBPS is a simple and widely available technique that can assess both LVEF and RVEF, and volumes with slight differences compared with CMR.

Full-volume three-dimensional echocardiographic assessment of left ventricular volume in healthy human fetuses

Objective To evaluate the feasibility and accuracy of full volume three-dimensional echocardiography(FV3DE) for quantifying the left ventricular volume in healthy fetuses. Methods Two hundred and twenty-two normal fetuses with gestational ages between 20 and 40 weeks were investigated. Full volume three-dimensional, two-dimensional (2D) and M-mode images were acquired by Philips iE33 system and data was processed offline using QLAB analysis software. Left ventricular end-systolic volume(ESV) and end-diastolic volume ( EDV) were measured using semiautomated border detection echocardiographic techniques, and compared with 2D biplane Simpson's method and M-mode Teichholz formula. Results Thirty-six fetuses with poor three-dimensional images and 5 twin pregnancies were excluded,the study subjects consisted of 181 singleton fetuses. The highest concordance correlation ( r =0.987 for EDV, r =0.988 for ESV) P <0. 001) was noted between 2D biplane Simpson's and FV3DE-derived values with a small bias (0. 332 ml for EDV,0. 135 ml for ESV). The Bland-Altman analysis showed that FV3DE had the best reproducibility than M-mode Teichholz formula and 2D biplane Simpson's method. Intra- and interobserver error of EDV were 2.85% and 4.22% for FV3DE,8.37% and 8.98% for 2D biplane Simpson's rule,30. 63% and 54. 70% for M-mode Teichholz formula. Intra- and interobserver error of ESV were 2.34% and 6. 14% for FV3DE,6. 26% and 11.7% for 2D biplane Simpson's rule,43.06% and 66.03% for M-mode Teichholz formula. Conclusions FV3DE is a novel technique that can be more accurate than conventional methods in quantification of fetal left ventricular volumes, it may provide new insight into evolving fetal cardiac structure and function. Key words: Ultrasonography,prenatal s Fetus; Cardiac volume,left

Three-Dimensional Ordered-Subset Expectation Maximization Iterative Protocol for Evaluation of Left Ventricular Volumes and Function by Quantitative Gated SPECT: A Dynamic Phantom Study

The purposes of this study were to characterize the performance of a 3-dimensional (3D) ordered-subset expectation maximization (OSEM) algorithm in the quantification of left ventricular (LV) function with (99m)Tc-labeled agent gated SPECT (G-SPECT), the QGS program, and a beating-heart phantom and to optimize the reconstruction parameters for clinical applications. A G-SPECT image of a dynamic heart phantom simulating the beating left ventricle was acquired. The exact volumes of the phantom were known and were as follows: end-diastolic volume (EDV) of 112 mL, end-systolic volume (ESV) of 37 mL, and stroke volume (SV) of 75 mL; these volumes produced an LV ejection fraction (LVEF) of 67%. Tomographic reconstructions were obtained after 10-20 iterations (I) with 4, 8, and 16 subsets (S) at full width at half maximum (FWHM) gaussian postprocessing filter cutoff values of 8-15 mm. The QGS program was used for quantitative measurements. Measured values ranged from 72 to 92 mL for EDV, from 18 to 32 mL for ESV, and from 54 to 63 mL for SV, and the calculated LVEF ranged from 65% to 76%. Overall, the combination of 10 I, 8 S, and a cutoff filter value of 10 mm produced the most accurate results. The plot of the measures with respect to the expectation maximization-equivalent iterations (I x S product) revealed a bell-shaped curve for the LV volumes and a reverse distribution for the LVEF, with the best results in the intermediate range. In particular, FWHM cutoff values exceeding 10 mm affected the estimation of the LV volumes. The QGS program is able to correctly calculate the LVEF when used in association with an optimized 3D OSEM algorithm (8 S, 10 I, and FWHM of 10 mm) but underestimates the LV volumes. However, various combinations of technical parameters, including a limited range of I and S (80-160 expectation maximization-equivalent iterations) and low cutoff values (< or =10 mm) for the gaussian postprocessing filter, produced results with similar accuracies and without clinically relevant differences in the LV volumes and the estimated LVEF.

Cardiac magnetic resonance versus transthoracic echocardiography for the assessment of cardiac volumes and regional function after myocardial infarction: an intrasubject comparison using simultaneous intrasubject recordings

BackgroundAlthough echocardiography is commonly used to evaluate cardiac function after MI, CMR may provide more accurate functional assessment but has not been adequately compared with echo. The primary study objective was to compare metrics of left ventricular volumes and global and regional function determined by cardiac magnetic resonance (CMR) and echocardiography (echo) in patients (pts) with recent myocardial infarction (MI).MethodsTo compare CMR with echo, 47 consecutive patients (pts 70% male; mean age = 66 ± 11 years) with MI >6 wks previously and scheduled for imaging evaluation were studied by both echo and CMR within 60 min of each other. Readers were blinded to pt information. Pearson's correlation coefficient, paired t-tests, and chi-square tests were used to compare CMR and echo measures. Further comparisons were made between pts and 30 normal controls for CMR and between pts and published normal ranges for echo.ResultsMeasures of volume and function correlated moderately well between CMR and echo (r = 0.54 to 0.75, all p < 0.001), but large and systematic differences were noted in absolute measurements. Echo underestimated left ventricular (LV) volumes (by 69 ml for end-diastolic, 35 ml for end-systolic volume, both p < 0.001), stroke volume (by 34 ml, p < 0.001), and LV ejection fraction (LVEF) (by 4 percentage point, p = 0.02). CMR was much more sensitive to detection of segmental wall motion abnormalities (p < 0.001). CMR comparisons with normal controls confirmed an increase in LV volumes, a decrease in LVEF, and preservation of stroke volume after MI.ConclusionThis intra subject comparison after MI found large, systematic differences between CMR and echo measures of volumes, LVEF, and wall motion abnormality despite moderate inter-modality correlations, with echo underestimating each metric. CMR also provided superior detection and quantification of segmental function after MI. Serial studies of LV function in individual patients should use the same modality.

Semi-automated quantification of left ventricular volumes and ejection fraction by real-time three-dimensional echocardiography

BackgroundRecent studies have shown that real-time three-dimensional (3D) echocardiography (RT3DE) gives more accurate and reproducible left ventricular (LV) volume and ejection fraction (EF) measurements than traditional two-dimensional methods. A new semi-automated tool (4DLVQ) for volume measurements in RT3DE has been developed. We sought to evaluate the accuracy and repeatability of this method compared to a 3D echo standard.MethodsLV end-diastolic volumes (EDV), end-systolic volumes (ESV), and EF measured using 4DLVQ were compared with a commercially available semi-automated analysis tool (TomTec 4D LV-Analysis ver. 2.2) in 35 patients. Repeated measurements were performed to investigate inter- and intra-observer variability.ResultsAverage analysis time of the new tool was 141s, significantly shorter than 261s using TomTec (p < 0.001). Bland Altman analysis revealed high agreement of measured EDV, ESV, and EF compared to TomTec (p = NS), with bias and 95% limits of agreement of 2.1 ± 21 ml, -0.88 ± 17 ml, and 1.6 ± 11% for EDV, ESV, and EF respectively. Intra-observer variability of 4DLVQ vs. TomTec was 7.5 ± 6.2 ml vs. 7.7 ± 7.3 ml for EDV, 5.5 ± 5.6 ml vs. 5.0 ± 5.9 ml for ESV, and 3.0 ± 2.7% vs. 2.1 ± 2.0% for EF (p = NS). The inter-observer variability of 4DLVQ vs. TomTec was 9.0 ± 5.9 ml vs. 17 ± 6.3 ml for EDV (p < 0.05), 5.0 ± 3.6 ml vs. 12 ± 7.7 ml for ESV (p < 0.05), and 2.7 ± 2.8% vs. 3.0 ± 2.1% for EF (p = NS).ConclusionIn conclusion, the new analysis tool gives rapid and reproducible measurements of LV volumes and EF, with good agreement compared to another RT3DE volume quantification tool.

Evaluation of Left Ventricular Function by 64-Multidetector Computed Tomography in Patients Undergoing Totally Endoscopic Coronary Artery Bypass Grafting

The goal of this study was to quantify left ventricular (LV) function with automated 3-dimensional volume segmentation by 64-slice computed tomography (CT) in patients undergoing totally endoscopic coronary artery bypass grafting (CABG). We used 64-multidetector CT coronary angiography to examine 63 patients with >70% coronary stenosis who were undergoing totally endoscopic CABG for single-vessel disease (left internal mammary artery to left anterior descending coronary artery) or multivessel disease with the da Vinci robotic surgical device (arrested heart approach). CT measurements were compared with cineventriculography results in 20 patients. The intraobserver variability values for the end-systolic volume (ESV) and the end-diastolic volume (EDV) were excellent (7.2% and 5.2%, respectively). Bland-Altman plots showed good upper and lower limits of agreement (ESV, +9% and -3.3%, respectively; EDV, +17% and -5.9%). Intraobserver variability for the LV ejection fraction (LVEF) was 4.8% with narrow limits of agreement (+7.8%, -2.2%). The mean postprocessing time was 6.5 minutes. Mean values (+/-SD) were 62.7% +/- 12% (range, 23%-86%) for LVEF, 98.4 mL +/- 29 mL for EDV, and 38.3 mL +/- 23 mL for ESV. The LVEF obtained via CT was moderately but significantly correlated with the invasive cineventriculogram (r = 0.51; P = .02; limits of agreement, +18.7% and -18.4%). . Through the use of automated LV volume segmentation, 64-slice CT permits fast quantification of LV function in patients with coronary artery disease undergoing totally endoscopic CABG grafting, enabling a comprehensive evaluation of coronary arteries and bypass grafts.

Intra-observer and interobserver reproducibility of right ventricle volumes, function and mass by cardiac magnetic resonance

Cardiac magnetic resonance (CMR) allows quick and non-invasive evaluation both of right ventricle (RV) volume and function, which are important in many heart diseases. We have evaluated CMR intra- and interobserver reproducibility in different conditions of RV dimension and function. We have analysed CMR exams of 45 subjects, randomly selected from our database according to RV end-diastolic volume (EDV; 15-subject groups with EDV < 25th, 25-75th and > 75th percentiles of a normal control population). Selected subjects were of both sexes (male/female 33/12) and of variable age (8-83 years) and body surface (0.9-2.3 m). RV end-diastolic and end-systolic volumes (ESV), ejection fraction (EF) and mass were blindly evaluated by two operators. Bland-Altman bias and coefficient of variability (CoV) were used to assess intra- and interobserver reproducibility. A wide range of EDV (range = 46-239 ml), ESV (20-129 ml) and EF (6-64%) was observed. The intra-observer bias was -5 ml for EDV, -2 ml for ESV, -1% for EF and 5 g for mass, with a CoV of 7-12%. The interobserver bias was 5 ml for EDV, 2 ml for ESV, 2% for EF and 6 g for mass, with a CoV of 8-13%. Analysis by tertiles showed EF assessment variability to be higher in the lower tertiles at intra-observer (P < 0.036) and, above all, at interobserver (P < 0.000) analysis. Mass assessment variability was higher in the upper tertile (P < 0.004) at intra-observer analysis. Intra- and interobserver reproducibility of RV parameters assessed by CMR are adequate in a wide range of RV dimensions and function. However, caution is required with respect to the significance of small changes of EF and mass in the case of poor function and hypertrophy of the RV, respectively.

New generation 3-dimensional echocardiography for left ventricular volumetric and functional measurements: Comparison with cardiac magnetic resonance

Non-invasive assessment of left ventricular (LV) structure and function is important in the evaluation of cardiac patients. This study was designed to test the accuracy and reproducibility of new generation 3-dimensional echocardiography (3DE) in measuring volumetric and functional LV indices as compared with current "gold standard" of non-invasive cardiac imaging, cardiac magnetic resonance (CMR). Sixty-four subjects with good acoustic windows, including 40 cardiac patients with LV ejection fraction (EF)<45%, 14 patients with EF>45% and 10 normal volunteers underwent 3DE using a commercially available Philips Sonos 7500 scanner equipped with a matrix phase-array x4 xMATRIX transducer, and CMR on a 1.5 T Signa CV/i scanner (GE Medical Systems). Volumetric assessment was performed with analytical 4D-LV-Analysis software (TomTec) for 3DE and MRI-Mass software (Medis) for CMR. We found no significant differences in LV end-diastolic volume (EDV), end-systolic volume (ESV) and EF with excellent correlations between the indices measured using 3DE and CMR (r=0.97, r=0.98, and r=0.94, respectively). Bland-Altman analysis showed bias of 7 ml for EDV, 3 ml for ESV and -1% for EF with 3DE with corresponding limits of agreement (2SD) of 28 ml, 22 ml and 10%, respectively. Intraobserver and interobserver variabilities were for EDV: 3% and 4% (3DE) vs 2% and 2% (CMR), for ESV: 3% and 6% (3DE) vs 2% and 3% (CMR), and for EF: 4% and 4% (3DE) vs 2% and 4% (CMR), respectively. New generation 3DE provides accurate and reproducible quantification of LV volumetric and functional data in subjects with good acoustic windows as compared with CMR.

Freehand Three‐Dimensional Echocardiography with Rotational Scanning for Measurements of Left Ventricular Volume and Ejection Fraction in Patients with Coronary Artery Disease

Measurement of left ventricular (LV) volumes and ejection fraction (EF) is important in managing patients with coronary artery disease (CAD). Introduction of free-hand three-dimensional echocardiography (3DE) system which is equipped with small magnetic tracking system and average rotational geometry for LV volumes may provide easy and accurate quantification of LV systolic function in CAD patients. To evaluate the feasibility and accuracy of LV volumes and EF measurement by free-hand 3DE with rotational geometry in patients with CAD. The study subjects consisted of consecutive 25 patients with CAD who were scheduled for quantitative gated single-photon emission computed tomography (QGS). LV end-diastolic volume (EDV), end-systolic volume (ESV), and EF were determined by conventional two-dimensional echocardiography (2DE), 3DE, and QGS. Three-dimensional echocardiography data acquisition and analysis were possible in 22 of 25 subjects (feasibility 88%). In this 3DE system, image acquisition time was 2 minutes, and 5 minutes were needed for off-line analysis of LV volumes and EF. Correlations and the limits of agreement between 3DE and QGS (r = 0.97, 0.0 +/- 9.1 ml for EDV, r = 0.99, 0.0 +/- 5.0 ml for ESV, and r = 0.97, 0.5 +/- 3.3% for EF, respectively) were superior to those between 2DE and QGS (r = 0.85, 12.6 +/- 26.8 ml for EDV, r = 0.85, 9.7 +/- 26.1 ml for ESV, and r = 0.90, -1.3 +/- 6.9% for EF, respectively). Inter- and intra-observer variabilities of 3DE were smaller than that of 2DE (5% vs 10%, 5% vs 10% for EDV, 6% vs 13%, 5% vs 9% for ESV, and 4% vs 11%, 4% vs 6% for EF, respectively). Three-dimensional echocardiography using magnetic tracking system and average rotational geometry offered a feasible and accurate method for quantification of LV volumes and EF in patients with CAD.