Segmentation of cervical intervertebral disks in videofluorography by CNN, multi-channelization and feature selection
Segmentation of cervical intervertebral disks in videofluorography by CNN, multi-channelization and feature selection
复制标题
通过 CNN、多通道化和特征选择对视频透视中的颈椎间盘进行分割
DOI：
10.1007/s11548-020-02145-8
复制
发表时间：
2020
影响因子：
3
通讯作者：
Hiroyuki Kudo
中科院分区：
文献类型：
--
作者：
Ayano Fujinaka;Kojiro Mekata;Hotaka Takizawa;Hiroyuki Kudo
目的吞咽困难是营养不良和吸入性肺炎的危险因素，对社会影响很大，因此有必要阐明吞咽困难的整个机制。在这项研究中，我们提出了一种分割方法的颈椎间盘（CID）在视频透视（VF），使用基于块的卷积神经网络（CNN），我们的多通道化（MC）方法和图像特征selection.MethodsTwenty图像滤波器分别适用于VF帧图像生成特征图像。通过将三个选定的特征图像设置为其红色、绿色和蓝色通道来生成一个彩色图像，称为多通道化图像。将基于块的CNN应用于MC图像，并通过基于像素的F-测度来评估CID的分割精度。这三个特征图像的组合是优化的模拟退火method.ResultsThe所提出的方法被应用到实际VF数据集组成的19名患者和39名健康参与者。当Sobel和形态学礼帽滤波器被选择在MC中的F-措施的分割精度为59.3%，而它是56.2%，当原始帧images.ConclusionThe实验结果表明，该方法是能够分割CID从实际VF，MC方法也能够提高分割精度约3%。在这项研究中，LeNet被用作CNN。我们未来的任务之一是使用其他CNN。
PurposeDysphagia has a large impact on the society because it is a risk factor of malnutrition and aspiration pneumonia, and therefore, it is necessary to elucidate the entire mechanism of dysphagia. In this study, we propose a segmentation method of cervical intervertebral disks (CIDs) in videofluorography (VF) by use of patch-based convolutional neural network (CNN), ourmulti-channelization(MC) method and image feature selection.MethodsTwenty image filters are individually applied to a VF frame image to generate feature images. One color image, called amulti-channelizedimage, is generated by setting three selected feature images to its red, green and blue channels. Patch-based CNN is applied to the MC image, and the segmentation accuracy of CIDs is evaluated by the pixel-basedF-measure. The combination of the three feature images is optimized by the simulated annealing method.ResultsThe proposed method was applied to actual VF dataset consisting of 19 patients and 39 healthy participants. The segmentation accuracy was 59.3% in theF-measure when Sobel and morphological top-hat filters were selected in MC, whereas it was 56.2% when original frame images were used.ConclusionThe experimental results demonstrated that the proposed method was able to segment CIDs from actual VF and also that the MC method was able to increase the segmentation accuracy by approximately 3%. In this study, LeNet was used as CNN. One of our future tasks is to use other CNNs.
登录
查看更多内容
DOI：
10.1007/978-3-319-66185-8
发表时间：
2017
期刊：
Medical Image Computing and Computer-Assisted Intervention − MICCAI 2017
影响因子：
--
作者：
Maxime Descoteaux;Lena Maier-Hein;A. Franz;P. Jannin;D. Collins;Simon Duchesne
通讯作者：
Maxime Descoteaux;Lena Maier-Hein;A. Franz;P. Jannin;D. Collins;Simon Duchesne
使用卷积神经网络从视频荧光图像中分割椎间盘
DOI：
10.1117/12.2521249
发表时间：
2019
期刊：
--
影响因子：
--
作者：
Ayano Fujinaka;Yukihoto Saito;Kojiro Mekata;H. Takizawa;H. Kudo
通讯作者：
H. Kudo
多通道化和CNN视频透视椎间盘分割的初步研究
DOI：
10.12792/icisip2019.050
发表时间：
2019
期刊：
Proceedings of The 7th International Conference on Intelligent Systems and Image Processing 2019
影响因子：
--
作者：
Ayano Fujinaka;Kojiro Mekata;H. Takizawa;H. Kudo
通讯作者：
H. Kudo
影响因子：
3.9
作者：
Lee, Jong Taek;Park, Eunhee;Jung, Tae-Du
通讯作者：
Jung, Tae-Du
影响因子：
2.6
作者：
Ryu, Ju Seok;Lee, Ji Hyun;Shin, Dong Ah
通讯作者：
Shin, Dong Ah