3D Ultrasound

3D ultrasound imaging aims to create volumetric representations of anatomical structures, overcoming the limitations of traditional 2D scans. Current research focuses on improving image quality through deep learning techniques like neural radiance fields and generative adversarial networks (GANs), as well as developing automated segmentation and registration methods using U-Net and other architectures for tasks such as placenta volume measurement and fetal pose estimation. These advancements are enhancing diagnostic accuracy and efficiency in various medical applications, particularly in obstetrics and gynecology, by providing more complete and readily interpretable anatomical information. The development of lightweight models is also a key focus to enable wider clinical adoption, especially in resource-constrained settings.