Author + information
- Anitha Parthiban,
- Rita France,
- Kimberley J. Reid,
- Girish Shirali and
- Sanket Shah
Background: Childhood cancer survivors (CCS) are at risk for ventricular dysfunction. Automated 3 D echocardiography (3DE) offers rapid, reproducible assessment of left ventricular (LV) volumes and ejection fraction (EF). We compared the accuracy and reproducibility of automated analysis (Auto) and manual tracking revision (MAN) to cardiac magnetic resonance imaging (CMR).
Methods: Enrolled CCS underwent 3DE and CMR on the same day. A single reader computed LV end diastolic and systolic volumes (EDV, ESV) and EF by automated software analysis of 3DE datasets (Tomtec 4DLV). A second reader repeated the measurements with manual adjustments of automated tracking. LV volumes and EF by CMR were computed using QMass MR v 6 (Medis). Percentage (%) error was calculated as difference between methods/mean.
Results: We enrolled 49 subjects, median age 17 years (range 10-25 years). Mean time taken for Auto EF was 0.39 ± 0.06 min vs 2.7± 0.8 min for manual EF (p<0.01). Tracking adjustments were needed for all datasets. The table shows % error in EF, EDV and ESV for inter-method and intra-method comparisons. Auto-CMR had higher % error when compared to MAN-CMR (p=0.01, p<0.01 and p=0.05 for EDV, ESV and EF respectively). Intra-reader % error was higher for MAN than Auto for all measurements (p<0.001).
Conclusions: While automated 3DE volumetrics is faster and more reproducible than manual 3DE, there are trade-offs in accuracy. Our results suggest the need for improvements in automated algorithms for pediatric applications.
Poster Hall, Hall C
Saturday, March 18, 2017, 3:45 p.m.-4:30 p.m.
Session Title: New Technologies in Echocardiography
Abstract Category: 28. Non Invasive Imaging: Echo
Presentation Number: 1245-220
- 2017 American College of Cardiology Foundation