Yu Wang1, Miao Fan2, Faiza Amber Siddiqui3, Meilian Wang4, Wei Sun1, Xue Sun1, Wenjia Lei1, Ying Zhang5. 1. Department of Sonography, Shengjing Hospital of China Medical University, Heping District, Shenyang, China. 2. Department of Radiology, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou, China. 3. Department of Entomology, The Pennsylvania State University, University Park, Pennsylvania. 4. Department of Entomology, The Pennsylvania State University, University Park, Pennsylvania; Department of Microbiology and Parasitology, College of Basic Medical Sciences, China Medical University, Heping District, Shenyang, China. 5. Department of Sonography, Shengjing Hospital of China Medical University, Heping District, Shenyang, China. Electronic address: baogoubei@hotmail.com.
Abstract
BACKGROUND: Obtaining an accurate diagnosis of fetal aortic arch anomalies is often difficult with traditional two-dimensional (2D) sonography. Thus, the aim of this study was to assess the value of three-dimensional (3D) sonography with spatiotemporal image correlation for diagnosing fetal aortic arch anomalies using a novel algorithm for 3D volume analysis. METHODS: Two-dimensional and 3D echocardiographic image data from 300 normal fetuses and 30 fetuses with aortic arch anomalies were retrospectively reviewed. Grayscale and high-definition flow imaging data were available for 2D echocardiography. Three-dimensional volumes were acquired in parasagittal views with high-definition flow imaging information. Volume postprocessing was performed using a novel algorithm to obtain 3D tomographic ultrasound imaging slices and color-rendered images. Detection of aortic arch positions, aberrant brachiocephalic patterns, and aortic arch anomalies was compared for 2D and 3D modalities. Postnatal echocardiography was used as the truth standard in assessing aortic anatomy. RESULTS: In total, four cases of double aortic arch, 21 cases of right aortic arch, one case of retroesophageal aortic arch, and four cases of left aortic arch with aberrant right subclavian arteries were included. Inter- and intraobserver agreement were excellent for 2D and 3D modalities. The 3D modality offered better sensitivity and accuracy compared with 2D imaging for the detection of brachiocephalic anomalies (P < .01) and arch anomalies (P < .01) but comparable sensitivity for arch position. There was no difference in specificity for both modalities. CONCLUSIONS: The proposed novel algorithm for volume postprocessing ensures that 3D reconstructed images are obtained with high repeatability. The addition of 3D spatiotemporal image correlation to fetal echocardiography may offer a more accurate diagnosis for arch anomalies.
BACKGROUND: Obtaining an accurate diagnosis of fetal aortic arch anomalies is often difficult with traditional two-dimensional (2D) sonography. Thus, the aim of this study was to assess the value of three-dimensional (3D) sonography with spatiotemporal image correlation for diagnosing fetal aortic arch anomalies using a novel algorithm for 3D volume analysis. METHODS: Two-dimensional and 3D echocardiographic image data from 300 normal fetuses and 30 fetuses with aortic arch anomalies were retrospectively reviewed. Grayscale and high-definition flow imaging data were available for 2D echocardiography. Three-dimensional volumes were acquired in parasagittal views with high-definition flow imaging information. Volume postprocessing was performed using a novel algorithm to obtain 3D tomographic ultrasound imaging slices and color-rendered images. Detection of aortic arch positions, aberrant brachiocephalic patterns, and aortic arch anomalies was compared for 2D and 3D modalities. Postnatal echocardiography was used as the truth standard in assessing aortic anatomy. RESULTS: In total, four cases of double aortic arch, 21 cases of right aortic arch, one case of retroesophageal aortic arch, and four cases of left aortic arch with aberrant right subclavian arteries were included. Inter- and intraobserver agreement were excellent for 2D and 3D modalities. The 3D modality offered better sensitivity and accuracy compared with 2D imaging for the detection of brachiocephalic anomalies (P < .01) and arch anomalies (P < .01) but comparable sensitivity for arch position. There was no difference in specificity for both modalities. CONCLUSIONS: The proposed novel algorithm for volume postprocessing ensures that 3D reconstructed images are obtained with high repeatability. The addition of 3D spatiotemporal image correlation to fetal echocardiography may offer a more accurate diagnosis for arch anomalies.