摘要
目的医学图像分割是医学图像处理的关键步骤。该文提出一种全卷积网络结构,实现X线图像中成像部位自动分割。方法鉴于卷积神经网络特征提取优势.设计由9个层级组成的全卷积网络进行医学图像分割。该网络采用多种尺寸卷积核来提取图像中多维图像特征,取消池化层以避免下采样过程中图像细节丢失。结果结合X线图像分割的特定场景进行实验:对比传统分割方法,该方法实现了更加精确的成像部位分割。结论全卷积网络能够提取有代表性的多维图像特征,避免下采样阶段图像细节丢失,实现X线图像中成像部位精确自动分割。
Objective Medical image segmentation is a key step in medical image processing.An architecture of fully convolutional networks was proposed to realize automatic segmentation of anatomical areas in X-ray images.Methods Enlightened by the advantages of convolutional neural networks on features extraction,fully convolutional networks consisting of 9 layers were designed to segment medical images.The networks used convolution kernels of various sizes to extract multi-dimensional image features in the images,meanwhile,eliminated pooling layers to avoid the loss of image details during downsampling procedures.Results The experiment was conducted in accordance with the specific scene of X-ray images segmentation.Compared with traditional segmentation methods,this approach achieved more accurate segmentation of anatomical areas.Conclusion Fully convolutional networks can extract representative and multidimensional features of medical images,avoid the loss of image details during downsampling procedures,and complete automatic segmentation of anatomical areas accurately in X-ray images.
引文
[1]施奕倩.影像组学的临床研究进展[J].实用放射学杂志,2017,33(10):1623-1626.
[2]代永亮,张元刚,方小东,等.计算机辅助的CT定量分析在早期单纯性肺挫伤研究中的价值[J].实用放射学杂志,2018,34(2):200-202.
[3]王章玉,杨翠微.基于改进型区域生长法的心脏三维建模的实现[J].中国医疗器械杂志,2014,38(5):315-317.
[4]MANSOOR A,BAGCI U,FOSTER B,et al.Segmentation and image analysis of abnormal lungs at CT:current approaches,challenges,and future trends[J].Radiographics,2015,35(4):1056-1076.
[5]王林,喻罡.基于扩散张量成像的婴幼儿大脑图像自动分割[J].中国医学影像技术,2017,33(8):1264-1268.
[6]SABOUR S,FROSST N,HINTON G E.Dynamic routing between capsules[C].Neural Inform Process Syst,2017:1-11.
[7]BOUVRIE J.Notes on convolutional neural networks[J].Neural Nets,2006,18(5):162-169.
[8]HINTON G E,OSINDERO S,TEH Y W.A fast learning algorithm for deep belief nets[J].Neural Comput,2006,18(7):1527-1554.
[9]KRIZHEVSKY A,SUTSKEVER I,HINTON G E.ImageNet classification with deep convolutional neural networks[C].Neural In form Process Syst,2012,60(2):1097-1105.
[10]SIMONYAN K,ZISSERMAN A.Very deep convolutional networks for large-scale image recognition[J].Comput Sci,2014.14(9):1556-1569.
[11]SZEGEDY C,LIU W,JIA Y,et al.Going deeper with convolutions[C].Proc IEEE Comput Soc Conf Comput Vis Pattern Recognit,2015,10(6):1-9.
[12]LONG J,SHELHAMER E,DARRELL T.Fully convolutional networks for semantic segmentation[C].Proc IEEE Comput Soc Conf Comput Vis Pattern Recognit,2015,79(10):3431-3440.
[13]CHEN L C,PAPANDREOU G,KOKKINOS I,et al.Semantic image segmentation with deep convolutional nets and fully connected CRFs[J].Comput Sci,2014,12(4):357-361.
[14]RONNEBERGER O,FISCHER P,BROX T.U-Net:convolutional networks for biomedical image segmentation[C].Med Imag Comput Comput Assist Interv,2015,93(51):234-241.
[15]郭磊,贺宏伟,王玉军,等.基于并行卷积神经网络的X线图像成像部位识别[J].中国医疗器械杂志,2018,42(2):92-94,98.
[16]ZHANG K,SONG H,ZHANG L.Active contours driven by local image fitting energy[J].Pattern Recognit,2010,43(4):1199-1206.
[17]WANG L,PAN C.Robust level set image segmentation via a local correntropy-based K-means clusterina[J].Pattern Recognit,2014.47(5):1917-19-25.