arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1807.05482cs.CV
收藏

基于块规范神经网络的近实时的海马分割

Near Real-time Hippocampus Segmentation Using Patch-based Canonical Neural Network

Imperial College London(帝国理工学院)

Zhongliu Xie, Duncan Gillies

摘要:

在过去的几十年中,最先进的医学图像分割在很大程度上依赖于信号处理范例,最值得注意的是基于配准的标签传播和成对补丁比较,尽管分割精度很高,但它们通常很慢。近年来,深度学习已经彻底改变了计算机视觉,其中许多实践优于现有技术,特别是关于图像分类的卷积神经网络(CNN)研究。深度CNN最近也开始应用于医学图像分割,但通常需要长时间的训练和苛刻的内存要求,取得的成功有限。我们提出了一个基于补丁的深度学习框架,该框架基于对经典神经网络模型的重新审视,并进行了大量的现代化,包括使用整流线性单元(ReLU)激活,dropout层,2.5D三平面补丁多路径设置。在使用来自ADNI数据库的100幅脑MR图像进行海马体分割的测试应用中,我们的方法在分割精度和速度方面都显著优于现有技术:在接近实时的【形】性能(<1s)上,中值Dice得分高达90.98%。

英文摘要:

Over the past decades, state-of-the-art medical image segmentation has heavily rested on signal processing paradigms, most notably registration-based label propagation and pair-wise patch comparison, which are generally slow despite a high segmentation accuracy. In recent years, deep learning has revolutionalized computer vision with many practices outperforming prior art, in particular the convolutional neural network (CNN) studies on image classification. Deep CNN has also started being applied to medical image segmentation lately, but generally involves long training and demanding memory requirements, achieving limited success. We propose a patch-based deep learning framework based on a revisit to the classic neural network model with substantial modernization, including the use of Rectified Linear Unit (ReLU) activation, dropout layers, 2.5D tri-planar patch multi-pathway settings. In a test application to hippocampus segmentation using 100 brain MR images from the ADNI database, our approach significantly outperformed prior art in terms of both segmentation accuracy and speed: scoring a median Dice score up to 90.98% on a near real-time performance (<1s).

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。

↑