Feasibility of left ventricular volume measurements by three-dimensional speckle tracking echocardiography depends on image quality and degree of left ventricular enlargement: validation study with cardiac magnetic resonance imaging.

BACKGROUND: Novel 3-dimensional echocardiography with speckle tracking imaging (3D-STE) may have advantages in assessing left ventricular (LV) volume through a cardiac cycle. The feasibility of 3D-STE may be affected by image quality and LV morphology.

METHODS AND RESULTS: We studied 64 patients (38 men, age 55±12 years) who underwent cardiac magnetic resonance imaging (CMRI) and 3D-STE on the same day. LV end-diastolic volume (EDV) and end-systolic volume (ESV) were measured by both modalities. Imaging qualities were quantified in each of 6 LV segments by an imaging quality score (IQS) of 1-3, and scores were averaged (mean IQS) at end-diastole and end-systole. Compared to CMRI, 3D-STE showed a tendency to underestimate LV volume measurements, but not significantly (EDV: bias=-18±37ml; ESV: bias=-10±34ml), and measurements correlated well with those by CMRI (EDV: R=0.80, ESV: R=0.86, ejection fraction: R=0.75, p<0.001). The absolute differences of LVEDV and ESV between 3D-STE and CMRI correlated significantly with mean IQS (LVEDV, R=-0.35, p=0.005; LVESV, R=-0.30, p=0.02). Based on the medium value of LVEDV by CMRI (127ml), subjects were classified into the small (<127ml) and large LVEDV (≧127ml) groups. In the large LVEDV group, mean IQS significantly correlated with the absolute differences of LVEDV (mean IQS, r=-0.45, p=0.01), despite no significant correlation in the small LVEDV group.

CONCLUSION: 3D-STE could measure LV volume as well as CMRI, however, its accuracy depends on the quality of the acquired image and particularly on enlargement of the left ventricle.