Validation of a novel automated border-detection algorithm for rapid and accurate quantitation of left ventricular volumes based on three-dimensional echocardiography

To assess the accuracy and reproducibility of a novel automated software for left ventricular (LV) volumes and ejection fraction (EF) measurements using real-time three-dimensional echocardiography (3DE). A total of 103 patients with a wide range of LV volumes were analyzed with both 4D AutoLVQ and...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:European heart journal cardiovascular imaging 2010-05, Vol.11 (4), p.359-368
Hauptverfasser: Muraru, Denisa, Badano, Luigi P, Piccoli, Gianluca, Gianfagna, Pasquale, Del Mestre, Lorenzo, Ermacora, Davide, Proclemer, Alessandro
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:To assess the accuracy and reproducibility of a novel automated software for left ventricular (LV) volumes and ejection fraction (EF) measurements using real-time three-dimensional echocardiography (3DE). A total of 103 patients with a wide range of LV volumes were analyzed with both 4D AutoLVQ and 4D TomTec software. In 23 patients, a side-by-side comparison of LV volume and EF measurements was done between 3DE, 2DE, and cardiac magnetic resonance (CMR). Excellent correlation was found between 4D AutoLVQ and 4D TomTec [r = 0.98 for end-diastolic volume (EDV), 0.99 for end-systolic volume (ESV), and 0.97 for EF, P < 0.0001], with small biases and narrow limits of agreement: EDV 5.2 mL (-14 to 25 mL), ESV 2.9 mL (-10 to 16 mL), EF -0.2% (-7 to 6%). Time of analysis was halved using 4D AutoLVQ with manual correction (1 min 52 s+/- 30 s) in comparison with 4D TomTec software (3 min 46 s +/- 1 min 24 s). Both softwares showed similar accuracy in comparison with CMR (4D AutoLVQ biases -11.0 mL, -9.1 mL, and 2.9%; 4D TomTec biases -8.3 mL, -7.4 mL, and 2.8% for EDV, ESV, and EF, respectively, P = NS for all) and good reproducibility. Novel 4D AutoLVQ software showed very good agreement with more time-consuming 4D TomTec software, having similar accuracy against CMR.
ISSN:2047-2404
1532-2114
2047-2412
DOI:10.1093/ejechocard/jep217