Introduction: Three-dimensional echocardiography (3DE) is now widely available, but is oft en underutilized in daily clinical practice. In theory, the ability of 3DE to display volumetric datasets should allow for increased accuracy in the identification of left ventricle wall motion abnormalities (LVWMAs), by encompassing the whole of the LV in the imaging sector. Sequential short-axis displays of the LV (I slice), comparable to those produced by cardiac MRI, can be obtained from an apical acquisition of a 3D dataset and may allow superior visualization of LVWMAs. We set out to assess the feasibility and the inter-observer variability of the ‘I slice’ mode for the assessment of LVWMAs. Methods: Five patients (pts) referred for assessment of LV systolic function had WMAs analyzed in 3 apical conventional 2D views (4 chamber, 2 chamber and long axis), using a standard 16 segment model and also in 3D I slice, using 9 sequential short axis slices (3 each for basal, mid and apical LV levels), reconstructed form 4 consecutive beats. All images were analyzed in a blinded manner by 9 observers (all EACVI- or BSEaccredited). Each segment received a WMA score, ranging from 0 to 4, according to current guidelines, and a WMA score was calculated in 2D and 3D I slice for each patient and each observer by adding the individual segment scores. We pooled the WMA scores of the 9 observers for each patient and compared them across patients and across methods. Results: Mean value for uninterpretable segments (no., %) reported for a patient was 5.8/4 for 2D and 2.8/1.9 for I slice. Mean value for the segments reported abnormal (no., %) was 97.4/67 for 2D and 85.2/59 for I slice. WMA score mean value and standard deviation were 32.78/2.72 (pt 1), 24.5/8.57 (pt 2), 37.8/6.95 (pt 3), 35.3/6.3 (pt 4), 33.7/7.5 (pt 5) and 32.6/4.9 (mean value) for 2D and 38.4/3.64 (pt 1), 21.1/5.7 (pt 2), 39.6/6.28 (pt 3), 28.3/5.7 (pt 4), 25.8/6.5 (pt 5) and 30.68/4.09 (mean value) for I slice. Th ere was a signifi cant correlation between mean WMS of the 9 observers by 2D and by 3D I slice (r = 0.7, p < 0.05). Conclusions: Using sequential short axis slices reconstructed from 3D datasets for wall motion scoring yields fewer uninterpretable segments than analysis of conventional 2D datasets, but comparable wall motion scores. The I-slice method merits corroboration with CMR in a future study
ISSN – online: 2734 – 6382
ISSN – print: 1220-658X
ISSN – print: 1220-658X
This work is licensed under a Creative Commons Attribution 4.0 International License.