Automatic 3D ultrasound segmentation of the first trimester placenta using deep learning

Abstract

First trimester placental volume measured with 3D ultrasound has been shown to be correlated to adverse pregnancy outcomes This could potentially be used as a screening test to predict the "at risk" pregnancy. However, manual segmentation whilst accurate is very time consuming. Semi-automated methods provide close agreement to manual segmentation but remain significantly operator dependant. To generate a screening tool fully automated placental segmentation is required. In this paper a previously published deep convolutional neural network, Deep Medic, was trained using the output of the semi-automated Random Walker method as the ground truth. A set of 300 ultrasound volumes was used to train, validate and test the neural network. Dice similarity coefficients from the neural network had a median value of 0.73. This work shows the feasibility for applying convolutional neural networks to automating segmentation of 3D ultrasound placental volume.

Fully-automated alignment of 3D fetal brain ultrasound to a canonical reference space using multi-task learning

“…In this paper, we present a single multi-task fully convolutional neural network which automatically locates and segments the fetal brain and eye sockets in 2D and 3D images. In the recent literature, 3D deep learning frameworks have proven successful in anatomical segmentation from volumetric ultrasound data ( Looney et al, 2017;Schmidt-Richberg et al, 2017;Yang et al, 2017 ). Our proposed model differs in that it makes simple, computationally inexpensive predictions from 2D slices and is capable of incorporating this information to estimate 3D brain orientation.…”

Fully automated, real-time 3D ultrasound segmentation to estimate first trimester placental volume using deep learning

“…Obtaining ground-truth data sets is challenging due to the laborious nature of data labeling, which typically is performed by clinicians experienced with the particular imaging modality. Efforts to segment the placenta using different fCNNs have been recently presented, but both used small data sets (5,6). A pilot study performed by the authors of this study using a different, simpler architecture and…”

“…only 300 cases obtained a DSC of 0.73 (5), while another demonstrated a DSC of 0.64 using 104 cases (6). While promising, what remains unclear is whether the DSC value, which is analogous to segmentation performance, is a result of the fCNN used or a reflection of the size the training set used.…”

Detection of Ovarian Tumors in Obstetric Ultrasound Imaging Using Logistic Regression Classifier With an Advanced Machine Learning Approach

“…In this paper, two parameters for comparison have been used, one classification accuracy, average accuracy and mean classification alert and the other mean average precision [25]. The precision-recall curve is a pixel dependent calculation using an uncertainty matrix for the evaluation of the performance [26] of the algorithm.…”