Fast and Accurate U-Net Model for Fetal Ultrasound Image Segmentation

Abstract

U-Net based algorithms, due to their complex computations, include limitations when they are used in clinical devices. In this paper, we addressed this problem through a novel U-Net based architecture that called fast and accurate U-Net for medical image segmentation task. The proposed fast and accurate U-Net model contains four tuned 2D-convolutional, 2D-transposed convolutional, and batch normalization layers as its main layers. There are four blocks in the encoder-decoder path. The results of our proposed architecture were evaluated using a prepared dataset for head circumference and abdominal circumference segmentation tasks, and a public dataset (HC18-Grand challenge dataset) for fetal head circumference measurement. The proposed fast network significantly improved the processing time in comparison with U-Net, dilated U-Net, R2U-Net, attention U-Net, and MFP U-Net. It took 0.47 seconds for segmenting a fetal abdominal image. In addition, over the prepared dataset using the proposed accurate model, Dice and Jaccard coefficients were 97.62% and 95.43% for fetal head segmentation, 95.07%, and 91.99% for fetal abdominal segmentation. Moreover, we have obtained the Dice and Jaccard coefficients of 97.45% and 95.00% using the public HC18-Grand challenge dataset. Based on the obtained results, we have concluded that a fine-tuned and a simple well-structured model used in clinical devices can outperform complex models.

Artificial intelligence in obstetric ultrasound: A scoping review

“…Five studies assessed deep learning for automated measurement of the FL only 59–63 and two studies assessed automated measurement of AC only 64,65 . There were four studies assessing correct detection of the fetal abdominal standard plane (FASP) using AI 66–69 and five studies assessed various combinations of two or more fetal biometry measurements 70–74 …”

“…Gofer were four studies assessing correct detection of the fetal abdominal standard plane (FASP) using AI [66][67][68][69] and five studies assessed various combinations of two or more fetal biometry measurements. [70][71][72][73][74] Ten studies focused on the use of AI in fetal cardiac imaging (Table 4). retrieves the right and left ventricular outflow tracts from a threedimensional (3D) volume of the fetal chest and found they were correctly identified 91.7% and 94.4% of the time, respectively.…”

Fully automated kidney image biomarker prediction in ultrasound scans using Fast-Unet++

“…Our previous paper proposed Fast-Unet 39 , which used convolution layers with a 5 × 5 kernel and 2 × 2 stride. Leaky Rectified Linear Units (Leaky ReLU) with 0.2 negative slope coefficient are set as the activation function in the encoder path and the ReLU activation function in the decoder path.…”

“…We have proposed a modification of Fast-Unet to overcome the problem. Fast-Unet is a high-performance novel CNN architecture that aims to segment fetal ultrasound images 39 . The key point in the network is using 2 × 2 stride in the convolution layers that downsamples the spatial resolution of feature maps, thus making the network needless to the pooling layer.…”

Automatic Detection of Secundum Atrial Septal Defect in Children Based on Color Doppler Echocardiographic Images Using Convolutional Neural Networks

“…The U-Net based architecture has also been widely applied in many segmentation tasks, such as liver ( 7 , 8 ), lung ( 9 ), tumor segmentation ( 10 , 11 ), and prostate ( 12 , 13 ). U-Net has also attracted many attentions in the field of ultrasonic images such as segmentation of ovary ( 14 ), fetal head ( 15 ), and breast ( 16 ). As for CHD diagnosis, it has also been reported that AI-based automatic auscultation may improve the accuracy of CHD screening ( 17 ).…”