OBJECTIVES: Conventional prenatal screening for congenital heart disease (CHD) involves a time-consuming and highly operator-dependent acquisition of the four-chamber view and outflow tracts. By acquiring the entire fetal heart instantaneously as a single volume, real-time three-dimensional echocardiography (RT3DE) may facilitate fetal cardiac screening. METHODS: Four reviewers, each experienced with fetal cardiac imaging, blindly and independently evaluated a single cardiac volume from each of 18 fetuses (11 normal, seven with CHD). Two-dimensional echocardiography served as the gold standard. Three-dimensional evaluation of each fetus included a series of volume acquisitions lasting 2-6 s each. A 'sweep volume' technique was developed to fit larger hearts into a single non-gated volume. RESULTS: RT3DE had a high sensitivity for detecting CHD (93%), with only a single case being missed by two observers. Specificity for CHD was low (45%), with a high rate of 'cannot determine' responses and false positive artifacts. CONCLUSIONS: These preliminary results suggest that RT3DE has the potential to function as a screening tool for fetal heart disease. However, artifacts must be recognized and minimized, resolution must improve, and substantial training will be necessary prior to widespread clinical use.
OBJECTIVES: Conventional prenatal screening for congenital heart disease (CHD) involves a time-consuming and highly operator-dependent acquisition of the four-chamber view and outflow tracts. By acquiring the entire fetal heart instantaneously as a single volume, real-time three-dimensional echocardiography (RT3DE) may facilitate fetal cardiac screening. METHODS: Four reviewers, each experienced with fetal cardiac imaging, blindly and independently evaluated a single cardiac volume from each of 18 fetuses (11 normal, seven with CHD). Two-dimensional echocardiography served as the gold standard. Three-dimensional evaluation of each fetus included a series of volume acquisitions lasting 2-6 s each. A 'sweep volume' technique was developed to fit larger hearts into a single non-gated volume. RESULTS: RT3DE had a high sensitivity for detecting CHD (93%), with only a single case being missed by two observers. Specificity for CHD was low (45%), with a high rate of 'cannot determine' responses and false positive artifacts. CONCLUSIONS: These preliminary results suggest that RT3DE has the potential to function as a screening tool for fetal heart disease. However, artifacts must be recognized and minimized, resolution must improve, and substantial training will be necessary prior to widespread clinical use.