Automated multivolume placental reconstruction using three-dimensional power Doppler ultrasound and infrared camera tracking.

Abstract

Objectives: Placental insufficiency contributes to many obstetric pathologies, but there is no bedside clinical tool to evaluate placental perfusion. We have developed a method to acquire multiple three-dimensional (3D) power Doppler (PD) ultrasound (US) volumes of placental vasculature, with infrared camera tracking of the precise spatial location of the transducer providing global coordinates. These volumes are reconstructed automatically ('stitched') into a model of the entire placenta. The purpose of this study was to evaluate the accuracy of automated reconstruction in an US phantom and to assess the feasibility of this technique in second-to-third-trimester human placentae.

Methods: A custom-designed acrylic phantom was constructed with dimensions mimicking a third-trimester placenta, containing 12 rectangular cuboid towers of various heights submersed in tissue-mimicking solution. Multiple overlapping 3D-US volumes of this phantom were acquired using three different insonation angles and infrared camera tracking. Data were transformed into a 3D cartesian volume and stitched automatically into six 3D-US volumes, each covering the entire phantom, for each of the three different insonation angles. Reconstruction accuracy was evaluated by calculating local distance error (assessment of towers in overlapping US volumes to determine accuracy of stitching) and global distance error (subtraction of true measurements in phantom model from corresponding measurements in stitched 3D-US volumes). A single-center, cross-sectional feasibility study was then conducted in women with an uncomplicated second-to-third-trimester singleton pregnancy, with data obtained using standardized ultrasound settings. Multiple 3D PD-US and grayscale volumes of the placentae were acquired with infrared camera-tracked coordinates. Volumes were stitched to create a model of placental vasculature, and these were assessed for quality and repeatability of volume measurement.

Results: Six entire phantom datasets were reconstructed at each of three insonation angles, giving a total of 18 extended phanom datasets. A median of nine 3D-US volumes required to reconstruct the entire phantom. Twelve towers per volume were assessed on three separate occasions, generating 648 datapoints. Of these datapoints, 67.1% were perfectly aligned. The mean local distance error was 2.92 (range, 0-25.51) mm. Measurements between towers of 120 distances in each stitched 3D-US volume (2160 distances in total) differed by an average of 1.51 (range, -4.78 to 4.23) mm from the true measurements in the phantom model. In the feasibility study, 17 participants were scanned, and 49 3D-US volume datasets acquired, with 92% reconstruction success per placental volume set and at least one complete volume being obtained per participant (100% participant achievability). The median volume acquisition and reconstruction time was 10 min. Reconstructed placental vasculature was assessed qualitatively to be present, continuous and detailed throughout. Volume measurement of entire segmented placentae was highly repeatable (intraclass correlation coefficient, 0.96 (95% CI, 0.89-0.99)).

Conclusion: We present an automated method to model the entire structure and vasculature of second-to-third-trimester placentae using multiple 3D PD-US volumes, with verified accuracy and clinical feasibility. This study builds a foundation on which to develop a practical screening tool for detecting placental insufficiency and has the potential for expansion to evaluation of adult organ perfusion. © 2024 The Author(s). Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of International Society of Ultrasound in Obstetrics and Gynecology.