OBJECTIVE: To compare the capability of three-dimensional (3D) reconstructed images produced by high-resolution episcopic microscopy (HREM) with that of in-vivo four-dimensional high-resolution transvaginal sonography (4D-HRTVS) to discern morphological features of the first-trimester human fetal heart. METHODS: This was a prospective study of fetal hearts between 9 and 14 weeks' gestation. For ex-vivo 3D analysis, 30 human fetal hearts (at 9 + 0 to 14 + 6 weeks) were retrieved from surgical terminations of pregnancy. The specimens were embedded in resin and episcopic ('block-face') imaging was used to obtain a digital volume dataset (HREM) using 3-micron slicing. 4D-HRTVS was performed in 28 separate pregnancies at 10 + 2 to 14 + 0 weeks using a Voluson E8 ultrasound machine with volumetric transvaginal RIC 6-12-MHz transducers. Heart volumes obtained by both methods were compared to assess their ability to demonstrate first-trimester cardiac morphology. Comparisons were made in the transverse and sagittal planes, and using volume rendering. RESULTS: All hearts were structurally normal, although abdominal situs was not examined in the isolated hearts that underwent HREM. 4D-HRTVS demonstrated each of the complete five transverse cardiac views in 32-86% of cases. HREM showed four features unique to the first-trimester human heart: prominent atrial appendages, spiral ventricular arrangement, prominent coronary arteries and thickened arterial walls. 4D-HRTVS could demonstrate the first two, but ultrasound resolution was too poor to quantify wall thickness and demonstrate coronary arteries in the 3-5-mm diameter heart. CONCLUSIONS: 4D-HRTVS showed limited morphological features of the first-trimester fetal heart compared with HREM. HREM provides a gold standard of ex-vivo imaging against which developments in ultrasound resolution could be compared.
OBJECTIVE: To compare the capability of three-dimensional (3D) reconstructed images produced by high-resolution episcopic microscopy (HREM) with that of in-vivo four-dimensional high-resolution transvaginal sonography (4D-HRTVS) to discern morphological features of the first-trimester human fetal heart. METHODS: This was a prospective study of fetal hearts between 9 and 14 weeks' gestation. For ex-vivo 3D analysis, 30 human fetal hearts (at 9 + 0 to 14 + 6 weeks) were retrieved from surgical terminations of pregnancy. The specimens were embedded in resin and episcopic ('block-face') imaging was used to obtain a digital volume dataset (HREM) using 3-micron slicing. 4D-HRTVS was performed in 28 separate pregnancies at 10 + 2 to 14 + 0 weeks using a Voluson E8 ultrasound machine with volumetric transvaginal RIC 6-12-MHz transducers. Heart volumes obtained by both methods were compared to assess their ability to demonstrate first-trimester cardiac morphology. Comparisons were made in the transverse and sagittal planes, and using volume rendering. RESULTS: All hearts were structurally normal, although abdominal situs was not examined in the isolated hearts that underwent HREM. 4D-HRTVS demonstrated each of the complete five transverse cardiac views in 32-86% of cases. HREM showed four features unique to the first-trimester human heart: prominent atrial appendages, spiral ventricular arrangement, prominent coronary arteries and thickened arterial walls. 4D-HRTVS could demonstrate the first two, but ultrasound resolution was too poor to quantify wall thickness and demonstrate coronary arteries in the 3-5-mm diameter heart. CONCLUSIONS: 4D-HRTVS showed limited morphological features of the first-trimester fetal heart compared with HREM. HREM provides a gold standard of ex-vivo imaging against which developments in ultrasound resolution could be compared.