BACKGROUND: The roles and accuracy of transesophageal (TEE) and real-time transthoracic (RT) three-dimensional echocardiography (3DE) are not clearly defined. The study purpose was to compare RT3DE and TEE-3D to surgical findings, and to assess the strengths and weaknesses of these techniques.

METHODS: Fourteen pre-operative patients over 11 kg with significant valve disease were included in the study. Each patient underwent pre-operative RT3DE and TEE two and 3D (Philips Medical Systems, SONOS 7500, X4 and TEE probes) studies. Median age was 7.9 years (range 2.25 – 14), median weight 29.9 kg (range 11 – 69.3). Diagnoses included: atrioventricular septal defect (3), mitral regurgitation or stenosis (6), tricuspid regurgitation (3), and aortic regurgitation (2). RT3DE data were acquired from apical and parasternal views for display, cropping and storage for offline review. TEE-3D data were acquired under anesthesia and stored for offline analysis (TomTec, Germany). Studies were performed, processed and analyzed in consensus by 2 observers. Surgical correlation was used as gold standard and assessed by reviewing the video-recording and surgical description. 3D images were scored (0: poor, 1: limited views, 2: fair, 3: good, 4: excellent) for image quality, valve morphology, subvalvar apparatus, colour jet information, and diagnostic findings.

RESULTS: Together, 3D images produced a new finding or markedly enhanced pathological views in 5 / 14 cases compared to TEE two-dimensional findings. Valve leaflet and commissure morphology was equally enhanced by either technique. TEE-3D image quality had no correlation to weight while RT3DE image quality was inversely correlated to weight (r = 0.6, p = 0.04).

CONCLUSION: Both modalities were able to delineate valve leaflet and commissure morphology from atrial and apical aspects. TEE-3D was superior in the visualisation of regurgitant colour jet number, shape, size and location, while RT3DE colour display suffered mostly from poor jet location. RT3DE is superior in imaging the subvalvar apparatus. Although TEE-3D requires anesthesia, a longer processing and acquisition time, echo windows do not limit the quality of the data sets.

DNC