A comparison of standard two-dimensional ultrasound to three-dimensional volume sonography for routine second-trimester fetal imaging

Assessment of Radiomics and Deep Learning for the Segmentation of Fetal and Maternal Anatomy in Magnetic Resonance Imaging and Ultrasound

Recent advances in fetal imaging open the door to enhanced detection of fetal disorders and computer-assisted surgical planning. However, precise segmentation of womb's tissues is challenging due to motion artifacts caused by fetal movements and maternal respiration during acquisition. This work aims to efficiently segment different intrauterine tissues in fetal magnetic resonance imaging (MRI) and 3D ultrasound (US). First, a large set of ninety-four radiomic features are extracted to characterize the mother uterus, placenta, umbilical cord, fetal lungs, and brain. The optimal features for each anatomy are identified using both K-best and Sequential Forward Feature Selection techniques. These features are then fed to a Support Vector Machine with instance balancing to accurately segment the intrauterine anatomies. To the best of our knowledge, this is the first time that "Radiomics" is expanded from classification tasks to segmentation purposes to deal with challenging fetal images. In addition, we evaluate several state-of-the-art deep learning-based segmentation approaches. Validation is extensively performed on a set of 60 axial MRI and 3D US images from pathological and clinical cases. Our results suggest that combining the selected 10 radiomic features per anatomy along with DeepLabV3+ or BiSeNet architectures for MRI, and PSPNet or Tiramisu for 3D US, can lead to the highest fetal / maternal tissue segmentation performance, robustness, informativeness, and heterogeneity. Therefore, this work opens new avenues for advancement of segmentation techniques and, in particular, for improved fetal surgical planning.

Towards automated extraction of 2D standard fetal head planes from 3D ultrasound acquisitions: A clinical evaluation and quality assessment comparison

Introduction

Clinical evaluation of deep learning (DL) tools is essential to compliment technical accuracy metrics. This study assessed the image quality of standard fetal head planes automatically-extracted from three-dimensional (3D) ultrasound fetal head volumes using a customised DL-algorithm.

Methods

Two observers retrospectively reviewed standard fetal head planes against pre-defined image quality criteria. Forty-eight images (29 transventricular, 19 transcerebellar) were selected from 91 transabdominal fetal scans (mean gestational age = 26 completed weeks, range = 20⁺⁵–32⁺³ weeks). Each had two-dimensional (2D) manually-acquired (2D-MA), 3D operator-selected (3D-OS) and 3D-DL automatically-acquired (3D-DL) images. The proportion of adequate images from each plane and modality, and the number of inadequate images per plane was compared for each method. Inter and intra-observer agreement of overall image quality was calculated.

Results

Sixty-seven percent of 3D-OS and 3D-DL transventricular planes were adequate quality. Forty-five percent of 3D-OS and 55% of 3D-DL transcerebellar planes were adequate.

Seventy-one percent of 3D-OS and 86% of 3D-DL transventricular planes failed with poor visualisation of intra-cranial structures. Eighty-six percent of 3D-OS and 80% of 3D-DL transcerebellar planes failed due to inadequate visualisation of cerebellar hemispheres. Image quality was significantly different between 2D and 3D, however, no significant difference between 3D-modalities was demonstrated (p < 0.005). Inter-observer agreement of transventricular plane adequacy was moderate for both 3D-modalities, and weak for transcerebellar planes.

Conclusion

The 3D-DL algorithm can automatically extract standard fetal head planes from 3D-head volumes of comparable quality to operator-selected planes. Image quality in 3D is inferior to corresponding 2D planes, likely due to limitations with 3D-technology and acquisition technique.

Implications for practice

Automated image extraction of standard planes from US-volumes could facilitate use of 3DUS in clinical practice, however image quality is dependent on the volume acquisition technique.

Computational modeling of the interactions between the maternal and fetal circulations in human pregnancy

In pregnancy, fetal growth is supported by its placenta. In turn, the placenta is nourished by maternal blood, delivered from the uterus, in which the vasculature is dramatically transformed to deliver this blood an ever increasing volume throughout gestation. A healthy pregnancy is thus dependent on the development of both the placental and maternal circulations, but also the interface where these physically separate circulations come in close proximity to exchange gases and nutrients between mum and baby. As the system continually evolves during pregnancy, our understanding of normal vascular anatomy, and how this impacts placental exchange function is limited. Understanding this is key to improve our ability to understand, predict, and detect pregnancy pathologies, but presents a number of challenges, due to the inaccessibility of the pregnant uterus to invasive measurements, and limitations in the resolution of imaging modalities safe for use in pregnancy. Computational approaches provide an opportunity to gain new insights into normal and abnormal pregnancy, by connecting observed anatomical changes from high-resolution imaging to function, and providing metrics that can be observed by routine clinical ultrasound. Such advanced modeling brings with it challenges to scale detailed anatomical models to reflect organ level function. This suggests pathways for future research to provide models that provide both physiological insights into pregnancy health, but also are simple enough to guide clinical focus. We the review evolution of computational approaches to understanding the physiology and pathophysiology of pregnancy in the uterus, placenta, and beyond focusing on both opportunities and challenges. This article is categorized under: Reproductive System Diseases >Computational Models.

Decision Fusion-Based Fetal Ultrasound Image Plane Classification Using Convolutional Neural Networks

Machine learning for ultrasound image analysis and interpretation can be helpful in automated image classification in large-scale retrospective analyses to objectively derive new indicators of abnormal fetal development that are embedded in ultrasound images. Current approaches to automatic classification are limited to the use of either image patches (cropped images) or the global (whole) image. As many fetal organs have similar visual features, cropped images can misclassify certain structures such as the kidneys and abdomen. Also, the whole image does not encode sufficient local information about structures to identify different structures in different locations. Here we propose a method to automatically classify 14 different fetal structures in 2-D fetal ultrasound images by fusing information from both cropped regions of fetal structures and the whole image. Our method trains two feature extractors by fine-tuning pre-trained convolutional neural networks with the whole ultrasound fetal images and the discriminant regions of the fetal structures found in the whole image. The novelty of our method is in integrating the classification decisions made from the global and local features without relying on priors. In addition, our method can use the classification outcome to localize the fetal structures in the image. Our experiments on a data set of 4074 2-D ultrasound images (training: 3109, test: 965) achieved a mean accuracy of 97.05%, mean precision of 76.47% and mean recall of 75.41%. The Cohen κ of 0.72 revealed the highest agreement between the ground truth and the proposed method. The superiority of the proposed method over the other non-fusion-based methods is statistically significant (p < 0.05). We found that our method is capable of predicting images without ultrasound scanner overlays with a mean accuracy of 92%. The proposed method can be leveraged to retrospectively classify any ultrasound images in clinical research.

Automated Fetal Head Detection and Circumference Estimation from Free-Hand Ultrasound Sweeps Using Deep Learning in Resource-Limited Countries

Ultrasound imaging remains out of reach for most pregnant women in developing countries because it requires a trained sonographer to acquire and interpret the images. We address this problem by presenting a system that can automatically estimate the fetal head circumference (HC) from data obtained with use of the obstetric sweep protocol (OSP). The OSP consists of multiple pre-defined sweeps with the ultrasound transducer over the abdomen of the pregnant woman. The OSP can be taught within a day to any health care worker without prior knowledge of ultrasound. An experienced sonographer acquired both the standard plane-to obtain the reference HC-and the OSP from 183 pregnant women in St. Luke's Hospital, Wolisso, Ethiopia. The OSP data, which will most likely not contain the standard plane, was used to automatically estimate HC using two fully convolutional neural networks. First, a VGG-Net-inspired network was trained to automatically detect the frames that contained the fetal head. Second, a U-net-inspired network was trained to automatically measure the HC for all frames in which the first network detected a fetal head. The HC was estimated from these frame measurements, and the curve of Hadlock was used to determine gestational age (GA). The results indicated that most automatically estimated GAs fell within the P2.5-P97.5 interval of the Hadlock curve compared with the GAs obtained from the reference HC, so it is possible to automatically estimate GA from OSP data. Our method therefore has potential application for providing maternal care in resource-constrained countries.

A Crossover Comparison of Standard and Telerobotic Approaches to Prenatal Sonography

OBJECTIVES

To determine the feasibility of a telerobotic approach to remotely perform prenatal sonographic examinations.

METHODS

Thirty participants were prospectively recruited. Participants underwent a limited examination (assessing biometry, placental location, and amniotic fluid; n = 20) or a detailed examination (biometry, placental location, amniotic fluid, and fetal anatomic survey; n = 10) performed with a conventional ultrasound system. This examination was followed by an equivalent examination performed with a telerobotic ultrasound system, which enabled sonographers to remotely control all ultrasound settings and fine movements of the ultrasound transducer from a distance. Telerobotic images were read independently from conventional images.

RESULTS

The mean gestational age ± SD of the 30 participants was 22.9 ± 5.3 weeks. Paired-sample t tests showed no statistically significant difference between conventional and telerobotic measurements of fetal head circumference, biparietal diameter, or single deepest vertical pocket of amniotic fluid; however, a small but statistically significant difference was observed in measurements of abdominal circumference and femur length (P < .05). Intraclass correlations showed excellent agreement (>0.90) between telerobotic and conventional measurements of all 4 biometric parameters. Of 21 fetal structures included in the anatomic survey, 80% of the structures attempted across all patients were sufficiently visualized by the telerobotic system (range, 57%-100% per patient). Ninety-seven percent of patients strongly or somewhat agreed that they would be willing to have another telerobotic examination in the future.

CONCLUSIONS

A telerobotic approach is feasible for remotely performing prenatal sonographic examinations. Telerobotic sonography (robotic telesonography) may allow for the development of satellite ultrasound clinics in rural, remote, or low-volume communities, thereby increasing access to prenatal imaging in underserved communities.

Segmentation and classification in MRI and US fetal imaging: Recent trends and future prospects

Fetal imaging is a burgeoning topic. New advancements in both magnetic resonance imaging and (3D) ultrasound currently allow doctors to diagnose fetal structural abnormalities such as those involved in twin-to-twin transfusion syndrome, gestational diabetes mellitus, pulmonary sequestration and hypoplasia, congenital heart disease, diaphragmatic hernia, ventriculomegaly, etc. Considering the continued breakthroughs in utero image analysis and (3D) reconstruction models, it is now possible to gain more insight into the ongoing development of the fetus. Best prenatal diagnosis performances rely on the conscious preparation of the clinicians in terms of fetal anatomy knowledge. Therefore, fetal imaging will likely span and increase its prevalence in the forthcoming years. This review covers state-of-the-art segmentation and classification methodologies for the whole fetus and, more specifically, the fetal brain, lungs, liver, heart and placenta in magnetic resonance imaging and (3D) ultrasound for the first time. Potential applications of the aforementioned methods into clinical settings are also inspected. Finally, improvements in existing approaches as well as most promising avenues to new areas of research are briefly outlined.