FPUS23
收藏资源简介:
FPUS23是一个专为评估胎儿超声成像而设计的胎儿幻影超声数据集,由维也纳工业大学创建。该数据集包含15,728张图像,用于训练四种不同的深度神经网络模型,以识别胎儿的正确诊断平面、方向及其解剖特征。数据集的创建过程涉及使用23周龄的胎儿幻影,避免了与医疗数据相关的法规限制,并通过科学家的专业标注确保了数据集的质量。FPUS23数据集的应用领域主要集中在提高临床工作流程和开发基于超声的胎儿监测平台,旨在解决胎儿超声图像的自动分析问题。
FPUS23 is a fetal phantom ultrasound dataset specifically designed for evaluating fetal ultrasound imaging, created by Vienna University of Technology. This dataset comprises 15,728 images, which are utilized to train four distinct deep neural network models to identify the correct diagnostic planes, orientations and anatomical features of fetuses. The development of this dataset employed a 23-week-old fetal phantom, thereby avoiding regulatory restrictions associated with medical data, and ensured its quality through professional annotations from scientists. The FPUS23 dataset is primarily applied to improving clinical workflows and developing ultrasound-based fetal monitoring platforms, with the aim of solving the problem of automated analysis of fetal ultrasound images.

- 1FPUS23: An Ultrasound Fetus Phantom Dataset with Deep Neural Network Evaluations for Fetus Orientations, Fetal Planes, and Anatomical Features维也纳工业大学 · 2023年



