Use of artificial intelligence and deep learning in fetal ultrasound imaging

Abstract

Deep learning is considered the leading artificial intelligence tool in image analysis in general. Deep-learning algorithms excel at image recognition, which makes them valuable in medical imaging. Obstetric ultrasound has become the gold standard imaging modality for detection and diagnosis of fetal malformations. However, ultrasound relies heavily on the operator's experience, making it unreliable in inexperienced hands. Several studies have proposed the use of deep-learning models as a tool to support sonographers, in an attempt to overcome these problems inherent to ultrasound. Deep learning has many clinical applications in the field of fetal imaging, including identification of normal and abnormal fetal anatomy and measurement of fetal biometry. In this Review, we provide a comprehensive explanation of the fundamentals of deep learning in fetal imaging, with particular focus on its clinical applicability.

Development of Artificial Intelligence Image Classification Models for Determination of Umbilical Cord Vascular Anomalies

“…The prior research apply AI to fetal ultrasound studies has been recently reviewed. 1,2 First, there has been segmentation or "coloring" of the different anatomies to aid in identification and quantification. 7,9 Next, AI and ML have been applied to measurement and fetal risk assessments.…”

“…1 An area of great potential is the application of AI to better diagnose birth defects both more precisely and sooner than conventional methods. 1,2 To further develop these methodologies, a collabora-tion between artificial intelligence scientists and maternal-fetal medicine specialists is crucial. In this retrospective study, blinded prenatal ultrasound images were used to develop a versatile set of protocols to test advanced computer modeling for classifying and identifying umbilical cord anomalies.…”

“…Current applications of AI/ML include: predicting fetal distress in pregnancy-induced hypertension, classification of amniotic fluid levels, fetal anatomy colorization, and the umbilical coiling index. 2,[6][7][8][9] A practical application of AI image classification should involve increased maternal-fetal risk and be of added benefit to sonographers and obstetricians. Therefore, a clinical exemplar using standard ultrasound images that correlates to increased maternal-fetal risks should be pursued.…”

Fetal magnetic resonance imaging artifacts: role of deep learning to improve imaging

“…Deep learning can acquire knowledge of image features (e.g. distributions and artifacts) during training 4,5 . Ultimately, such algorithms should be refined before clinical use so that we can be certain that anomalies are not masked.…”

Machine Vision and Image Analysis in Anesthesia: Narrative Review and Future Prospects

“…5,6 The use of MV to augment the use of ultrasound (US) imaging is an active area of research, especially in fetal medicine. 7 MV has the potential to aid anesthesiologists in multiple areas of perioperative care (Figure 2). Imaging data can also be combined with nonimage data, such as patient demographics, medical history, surgical factors, to create hybrid ML models.…”