New speckle-tracking algorithm for right ventricular volume analysis from three-dimensional echocardiographic data sets: validation with cardiac magnetic resonance and comparison with the previous analysis tool

    loading  Checking for direct PDF access through Ovid

Abstract

Aims

(i) To validate a new software for right ventricular (RV) analysis by 3D echocardiography (3DE) against cardiac magnetic resonance (CMR); (ii) to assess the accuracy of different measurement approaches; and (iii) to explore any benefits vs. the previous software.

Methods and results

We prospectively studied with 3DE and CMR 47 patients (14–82 years, 28 men) having a wide range of RV end-diastolic volumes (EDV 82–354 mL at CMR) and ejection fractions (EF 34–81%). Multi-beat RV 3DE data sets were independently analysed with the new software using both automated and manual editing options, as well as with the previous software. RV volume reproducibility was tested in 15 random patients. RV volumes and EF measurements by the new software had an excellent accuracy (bias ± SD: −15 ± 24 mL for EDV; 1.4 ± 4.9% for EF) and reproducibility compared with CMR, provided that the RV borders automatically tracked by software were systematically edited by operator. The automated analysis option underestimated the EDV, overestimated the ESV, and largely underestimated the EF (bias ± SD: −17 ± 10%). RV volumes measured with the new software using manual editing showed similar accuracy, but lower inter-observer variability and shorter analysis time (3–5′) in comparison with the previous software.

Conclusion

Novel vendor-independent 3DE software enables an accurate, reproducible and faster quantitation of RV volumes and ejection fraction. Rather than optional, systematic verification of border tracking quality and manual editing are mandatory to ensure accurate 3DE measurements. These findings are relevant for echocardiography laboratories aiming to implement 3DE for RV analysis for both research and clinical purposes.

Related Topics

    loading  Loading Related Articles