arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2009.00872eess.IVcs.CVcs.LG
收藏

使用多尺度特征提取的高效、高性能胰腺分割

Efficient, high-performance pancreatic segmentation using multi-scale feature extraction

Technical University of Munich(慕尼黑工业大学);Institute for Artificial Intelligence in Medicine and Healthcare, Technical University of Munich(慕尼黑工业大学医学与健康人工智能研究所);OpenMined;Imperial College London(帝国理工学院)

Moritz Knolle, Georgios Kaissis, Friederike Jungmann, Sebastian Ziegelmayer, Daniel Sasse, Marcus Makowski, Daniel Rueckert, Rickmer Braren

摘要:

为了使基于人工智能的图像分析方法达到临床应用,开发高性能算法至关重要。例如,现有的基于自然图像的分割算法在其参数使用方面既不高效,也不针对医学成像进行优化。在这里,我们提出了MoNet,一种高度优化的基于神经网络的胰腺分割算法,专注于通过有效的多尺度图像特征利用来实现高性能。

英文摘要:

For artificial intelligence-based image analysis methods to reach clinical applicability, the development of high-performance algorithms is crucial. For example, existent segmentation algorithms based on natural images are neither efficient in their parameter use nor optimized for medical imaging. Here we present MoNet, a highly optimized neural-network-based pancreatic segmentation algorithm focused on achieving high performance by efficient multi-scale image feature utilization.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑