Abstract:A novel method is developed for the fetal abdominal contour extraction and measurement in ultrasound images. Fetal abdominal circumference (AC) is one of the standardized measurements in the antepartum ultrasound monitoring. Among several standardized measurements, AC is best correlated with fetal growth but is also the most difficult to be accurately measured. To overcome the difficulties in the abdominal contour extraction, the proposed method is a four-step procedure that integrates several image segmentation techniques. The proposed method is able to make the best use of the strength of different segmentation algorithms, while avoiding their deficiencies. An enhanced instantaneous coefficient of variation (ICOV) edge detector is first developed to detect edges of the abdominal contour and alleviate the effects of most speckle noise. Then, the Fuzzy C-Means clustering is employed to distinguish salient edges attributable to the abdominal contour from weak edges due to the other texture. Subsequently, the iterative Hough transform is applied to determine an elliptical contour and obtain an initial estimation of the AC. Finally, the gradient vector field (GVF) snake adapts the initial ellipse to the real edges of the abdominal contour. Quantitative validation of the proposed method on synthetic images under different imaging conditions achieves satisfactory segmentation accuracy (98.78% ± 0.16%). Experiments on 150 clinical images are carried out in three aspects: comparisons between inter-observer and inter-run variation, the fitness analysis between the automatically detected ellipse and the manual delineation, and the accuracy comparisons between automatic measurements and manual measurements in estimation of fetal weight (EFW). Experimental results show that the proposed method can provide consistent and accurate measurements. The reductions of the mean absolute difference and the standard deviation of EFW based on automatic measurements are about 1.2% and 2.1%, respectively, which indicate its potential in clinical antepartum monitoring application. (E-mail: yywang@fudan.edu.cn)

Fetal Lung Segmentation Using Texture-Based Boundary Enhancement and Active Contour Models

Automated 3D Fetal Brain Segmentation Using an Optimized Deep Learning Approach

Automatic Segmentation of a Fetal Echocardiogram Using Modified Active Appearance Models and Sparse Representation.

Simultaneous Segmentation of Fetal Hearts and Lungs for Medical Ultrasound Images Via an Efficient Multi-scale Model Integrated with Attention Mechanism

Segmentation of Ultrasound Fetal Images

Ultrasonic Evaluation of Fetal Lung Development Using deep learning with Graph

Joint Segmentation and Landmark Localization of Fetal Femur in Ultrasound Volumes

Fetal abdominal contour extraction and measurement in ultrasound images.

Segmentation of four-chamber view images in fetal ultrasound exams using a novel deep learning model ensemble method

Learning to segment fetal brain tissue from noisy annotations

Hybrid attention for automatic segmentation of whole fetal head in prenatal ultrasound volumes

Automated Detection of the Fetal Brain Midline and the Fetal Skull in 3-D Ultrasound Images Using the Modified Hough Transform

Detection of fetal facial anatomy in standard ultrasonographic sections based on real‐time target detection network

Deep Learning-Based Multiclass Brain Tissue Segmentation in Fetal MRIs

Fetal ultrasound image segmentation system and its use in fetal weight estimation

Segmentation of Fetal Left Ventricle in Echocardiographic Sequences Based on Dynamic Convolutional Neural Networks

Fetal Ultrasound Image Segmentation and Automatic Diameter and Length Measurement

A deep learning framework for identifying and segmenting three vessels in fetal heart ultrasound images

Automatic quality assessment for 2D fetal sonographic standard plane based on multi-task learning

Automatic Segmentation of the Fetal Cerebellum on 3-D Ultrasound Volumes Using Directional Phase Symmetry

Active Cardiac Model and Its Application on Structure Detection from Early Fetal Ultrasound Sequences