Feasibility of image registration and fusion for evaluation of structure and perfusion of the entire second trimester placenta by three-dimensional power Doppler ultrasound. (May 2020)

• Record Type: Journal Article
• Authors: Cheung, Winnie; Stevenson, Gordon N.; de Melo Tavares Ferreira, Ana Elizabeth Gomes; Alphonse, Jennifer; Welsh, Alec W.
• Is Part Of: Placenta. Volume 94(2020)
• Page Start: 13
• Page End: 19
• Publication Date: 2020-05
• Subjects: Imaging -- Three-dimensional -- Pregnancy trimester -- Second -- Placental circulation -- Placenta -- Diagnostic imaging* -- Ultrasonography -- Doppler; 3D Three Dimensional -- CoV Coefficient of Variation -- DSC Dice Similarity Coefficient -- FMBV Fractional Moving Blood Volume -- ICC Intra-class Correlation Coefficients -- IVS Inter-villous Space -- PD Power Doppler -- NIFTI Neuroimaging Informatics Technology Initiative -- UPI Utero-Placental Interface; Placenta -- Periodicals; Reproduction -- Periodicals; Placenta -- Periodicals; Placenta -- Périodiques; Reproduction -- Périodiques; 612.63
• DOI: 10.1016/j.placenta.2020.03.005
• Languages: English
• ISSNs: 0143-4004

Abstract

Abstract: Background: Placental perfusion can be evaluated by 3D power Doppler ultrasound (3D PD-US), particularly using the validated tool 3D Fractional Moving Blood Volume (3D-FMBV); however regional variability and size limitations beyond the first trimester mean that multiple 3D PD-US volumes are required to evaluate the whole organ. Purpose: We assessed the feasibility of manual offline stitching of second trimester 3D PD-US volumes of the placenta to assess whole organ perfusion using 3D-FMBV. Materials and methods: This was a single-centre, prospective, observational cohort study of 36 normal second trimester singleton pregnancies with anterior placentas. 3D PD-US placental volumes were manually segmented offline and stitched together by rigid registration using manually selected, pair-wise coordinates. Data acquisition and offline volume segmentation and stitching were triplicated by a single observer with Dice similarity coefficient (DSC) and Hausdorff distance used to assess consistency. Intraclass correlation coefficient (ICC) was used to assess intra-observer repeatability of 3D-FMBV and placental volume. Results: Acquisition and stitching success were 94% and 88%, respectively. Median time for acquisition, segmentation and stitching were 13 min, 40 min and 95 min, respectively. Median intra-observer DSCs were 0.94 and 0.88, and Hausdorff distances were 11.85 mm and 36.6 mm, for segmentations and stitching, respectively. Conclusion: 3D-ultrasound volume stitching of the placenta is technically feasible. Intra-observer repeatability was good to excellent for all measured parameters. This work demonstrates technical feasibility; further studies may provide the basis of an in-vivo assessment tool to measure the placenta in mid-to late pregnancy. Highlights: Article Highlights. Demonstrates use of image processing to show whole placental perfusion in mid-late pregnancy using 3D ultrasound (3D-US). The technique shows good intra-observer repeatability for all measured parameters (ICCs > 0.82). Triplication of labelling of the placenta led to excellent repeatability when measured by similarity metrics.