摘要
提出了一种基于全卷积神经网络(FCN)的左心室分割算法,该算法首先对图像进行三维块匹配(BM3D)去噪,然后在FCN模型加入批正则归一化处理,使得分割准确性得到提高,泛化能力加强。本文采用Sunnybrook数据集进行实验,验证了本分割算法的有效性。
In this paper,a left ventricle segmentation algorithm based on full convolutional neural network(FCN)is proposed.The algorithm firstly performs BM3D image denoising,and then adds batch regularization normalization processing to the FCN model to improve the segmentation accuracy and enhance the generalization ability.In this paper,experiments are carried out with Sunnybrook data set to verify the effectiveness of the proposed segmentation algorithm.
作者
徐春雨
王豫峰
刘英田
XU Chun-yu;WANG Yu-feng;LIU Ying-tian(College of Information Engineering,Nanyang Institute of Technology,Nanyang 473004,China;School of Computer and Software,Nanyang Institute of Technology,Nanyang 473004,China)
出处
《南阳理工学院学报》
2020年第6期53-57,共5页
Journal of Nanyang Institute of Technology
基金
河南省科技厅科技攻关项目(202102310199,192102310477)
河南省高等学校重点科研项目(20A520030,19A520029,19B520017)
南阳理工学院交叉科学研究项目。
