Research On Deep Learning Applied To Ultrasound Fetal Head Circumference Automatic Measurement

Fetal head circumference is an important biological index of fetus.By measuring fetal head circumference,obstetricians and gynecologists can predict the gestational age and due date of pregnant women,evaluate the development of fetus and the delivery mode of pregnant women.Accurate measurement of fetal head circumference requires sonographers with rich clinical experience,and the difference of clinical experience between sonographers will lead to certain deviation of the measurement results.Computer-assisted measurement of fetal head circumference can reduce the measurement error between different sonographers,reduce the working pressure,and improve the working efficiency of doctors Therefore,the study of the fetal head circumference automatic measurement algorithm is of great clinical significance.In clinical practice and the research of automatic measurement algorithm of head circumference,the measurement of head circumference is based on the assumption that the fetal head is oval in shape,and the circumference of the elliptic curve is the fetal head circumference that needs to be measured.The automatic measurement algorithm of fetal head circumference has three main steps:(1)The detection of fetal head edge in ultrasonic image;(2)Head edge elliptic curve obtained by elliptic fitting method;(3)Compute fetal head circumference.Fetal head edge detection is necessary for the automatic measurement of fetal head circumference.Ultrasound fetal head image boundary is fuzzy,and the gray scale of fetal head is similar to the mother’s abdominal tissue,especially in the first trimester.Ultrasound shadow leads to the loss of head edge and incomplete fetal head in the image,which brings certain difficulties in detecting the complete fetal head edge and fit head ellipse.The structures of the amniotic fluid and uterine wall are similar to the head texture and gray scale,often leading to misclassification of this part as fetal head.All these factors result in challenges to ultrasound fetal head edge detection.This paper based on the medical image segmentation network named U-Net.In order to improve the performance of U-Net,we adjust the architecture of U-Net,and use the improved U-Net to segment fetal head,then extract the edge of the head and calculate fetal head circumference.Semantic gap between different level of features which extract from U-Net is very large.In order to solve this problem,A new network named UNet++is proposed by scientist,UNet++network redesign the encoder decoder connection mode to reduce semantic gap between different level of features.Further on the basis of UNet++neural network structure,we fuse features of output layer by concatenation and further extract fused features.The improved model is named Fusion UNet++.To prevent overfitting,we introduce spatial dropout after each convolutional layer instead of standard dropout,which extends the dropout value across the entire feature map.The idea of fetal head circumference measurement is as follows:First,we use Fusion UNet++to learn the features of 2D ultrasound fetal head image and obtain the semantic segmentation result of the fetal head by using fetal head probability map.Second,on the basis of the image segmentation result,we extract the fetal head edge by using an edge detection algorithm and use the direct least square ellipse fitting method to fit the head contour.Finally,the fetal head circumference can be calculated using the ellipse circumference formula.The data in this paper are come from the open dataset of the automated measurement of fetal head circumference of the 2D ultrasound image named HC18 on Grand Challenges in Biomedical Image Analysis,which contains the first,second,and third trimester images of fetal heads.All fetal head images are the standard plane of measuring fetal head circumference.In the HC18 dataset,999 2D ultrasound images have annotations of fetal head circumference in the train set,and 335 2D ultrasound fetal head images have no annotations in the test set.We use the train set to train the convolutional neural network and submit the predicted results of the test set to participate in the model evaluation on HC18,Grand Challenges We use the Dice coefficient,Hausdorff distance(HD),and absolute difference(AD)as assessment indexes to evaluate the proposed method quantitatively.With improved-Net,for the dataset of fetal head images for all three trimesters,the Dice coefficient of the fetal head segmentation is 97.92%,the HD is 1.29±0.80 mm,and the AD of the fetal head circumference measurement is 1.93±1.91 mm.Our model could accurately locate the fetal head and identify the head area.With the proposed method fusion UNet++,for the dataset of fetal head images for all three trimesters,the Dice coefficient of the fetal head segmentation is 98.06%,the HD is 1.21±0.69 mm,and the AD of the fetal head circumference measurement is 1.84±1.73 mm.On the basis of the results presented in the open test set,Fusion UNet++achieves good results in the segmentation of the region of interest and the measurement of the head circumference.In comparison with the traditional and machine learning methods,the proposed method can effectively overcome the interference of fuzzy boundary and lack of edge and can accurately segment the fetal head region.In comparison with existing neural network methods,the proposed method surpasses the other methods in the second trimester of pregnancy in fetal head segmentation and head circumference measurement.The proposed method achieves the state-of-the-art results of fetal head segmentation.