Location Sensitive Deep Convolutional Neural Networks for Segmentation of White Matter Hyperintensities.

Location Sensitive Deep Convolutional Neural Networks for Segmentation of White Matter Hyperintensities.
复制标题

DOI:
10.1038/s41598-017-05300-5
复制
发表时间:
2017-07-11
期刊:
影响因子:
4.6
通讯作者:
Platel B
Platel B
中科院分区:
综合性期刊3区
文献类型:
--
作者:
Ghafoorian M;Karssemeijer N;Heskes T;van Uden IWM;Sanchez CI;Litjens G;de Leeuw FE;van Ginneken B;Marchiori E;Platel B

文献摘要

参考文献

被引文献

相似文献

成像特征的解剖位置对于许多医疗任务中的准确诊断至关重要。卷积神经网络 (CNN) 在计算机视觉领域取得了巨大成功,但它们缺乏将解剖位置纳入决策过程的天然能力,从而阻碍了某些医学图像分析任务的成功。在本文中,为了将解剖位置信息集成到网络中,我们提出了几种深度 CNN 架构,它们在训练时考虑多尺度补丁或采用显式位置特征。我们应用并比较了所提出的架构,用于在大型数据集上对大脑 MR 图像中的白质高信号进行分割。因此,我们观察到,整合位置信息的 CNN 大大优于具有手工特征的传统分割方法以及不整合位置信息的 CNN。在 50 次扫描的测试集中,我们网络的最佳配置获得了 0.792 的 Dice 分数,而独立人类观察者的 Dice 分数为 0.805。机器和独立人类观察者的性能水平没有统计学上的显着差异(p值 = 0.06)。
The anatomical location of imaging features is of crucial importance for accurate diagnosis in many medical tasks. Convolutional neural networks (CNN) have had huge successes in computer vision, but they lack the natural ability to incorporate the anatomical location in their decision making process, hindering success in some medical image analysis tasks. In this paper, to integrate the anatomical location information into the network, we propose several deep CNN architectures that consider multi-scale patches or take explicit location features while training. We apply and compare the proposed architectures for segmentation of white matter hyperintensities in brain MR images on a large dataset. As a result, we observe that the CNNs that incorporate location information substantially outperform a conventional segmentation method with handcrafted features as well as CNNs that do not integrate location information. On a test set of 50 scans, the best configuration of our networks obtained a Dice score of 0.792, compared to 0.805 for an independent human observer. Performance levels of the machine and the independent human observer were not statistically significantly different (p-value = 0.06).
DOI: 10.1109/tmi.2016.2528821
发表时间: 2016-05-01
影响因子: 10.6
作者:
Brosch, Tom;Tang, Lisa Y. W.;Tam, Roger
通讯作者: Tam, Roger
DOI: 10.1016/j.neuroimage.2005.06.061
发表时间: 2005-11-15
期刊: NEUROIMAGE
影响因子: 5.7
作者:
Admiraal-Behloul, F;van den Heuvel, DMJ;Relber, JHC
通讯作者: Relber, JHC
DOI: 10.1016/j.neunet.2012.02.023
发表时间: 2012-08-01
期刊: NEURAL NETWORKS
影响因子: 7.8
作者:
Ciresan, Dan;Meier, Ueli;Schmidhuber, Juergen
通讯作者: Schmidhuber, Juergen
DOI: 10.1007/s12021-015-9260-y
发表时间: 2015-07
期刊: NEUROINFORMATICS
影响因子: 3
作者:
Caligiuri, Maria Eugenia;Perrotta, Paolo;Augimeri, Antonio;Rocca, Federico;Quattrone, Aldo;Cherubini, Andrea
通讯作者: Cherubini, Andrea
DOI: 10.1016/j.neuroimage.2003.10.012
发表时间: 2004-03-01
期刊: NEUROIMAGE
影响因子: 5.7
作者:
Anbeek, P;Vincken, KL;van der Grond, J
通讯作者: van der Grond, J