| 改进的全卷积神经网络的脑肿瘤图像分割 |
| |
| 引用本文： | 邢波涛,李锵,关欣.改进的全卷积神经网络的脑肿瘤图像分割[J].信号处理,2018,34(8):911-922. |
| |
| 作者姓名： | 邢波涛 李锵 关欣 |
| |
| 作者单位： | 天津大学微电子学院 |
| |
| 基金项目： | 国家自然科学基金（61471263）；天津市自然科学基金（16JCZDJC31100）
|
| |
| 摘 要： | 针对现有机器学习算法分割脑肿瘤图像精度不高的问题，提出一种基于改进的全卷积神经网络的脑肿瘤图像分割算法。算法首先将FLAIR、T2和T1C三种模态的MR脑肿瘤图像进行灰度归一化,随后利用灰度图像融合技术得到肿瘤信息更加全面的预处理图像；然后采用融合三次脑肿瘤特征信息的改进全卷积神经网络对预处理图像进行粗分割，并且在每个卷积层后加入批量正则化层以加快网络训练的收敛速度，提高训练模型精度；最后融合全连接条件随机场细化粗分割结果中的脑肿瘤边界。实验结果表明，相较于传统的卷积神经网络脑肿瘤图像分割算法，本算法在分割精度和稳定性上有了较大提升，平均Dice可达91.29%，实时性较好，利用训练模型平均1s内可完成单张脑肿瘤图像的分割。
|
| 关 键 词： | 脑肿瘤图像分割 全卷积神经网络 条件随机场 |
| 收稿时间： | 2018-05-14 |
A brain tumor image segmentation method based on improved fully convolutional neural network |
| |
| Affiliation: | School of Microelectronics, Tianjin University |
| |
| Abstract: | The segmentation accuracy of the existing machine learning algorithms for brain tumor image segmentation is not high. So a brain tumor image segmentation method based on improved fully convolutional neural network (FCNN) is proposed. Firstly, the MR brain tumor images of FLAIR T2 and T1C are gray normalized. And then gray image fusion method is used to get more brain tumor feature information. Furthermore, brain tumor image is segmented cursorily with the improved fully convolutional neural network which fusing the tertiary pooled feature information. In order to speed up the convergence degree and improve the accuracy of the training model, batch normalization (BN) layer is added after each convolutional layer. Finally, the conditional random fields (CRF) is integrated to the fully convolutional neural network to fine the segmentation result. Compared with the traditional MRI brain tumor segmentation methods of convolutional neural network (CNN), the experimental results show that the segmentation accuracy and stability has been greatly improved. Average Dice can be up to 91.29%. And the real-time performance of the proposed method is good. A brain tumor image can be segmented with the training model within an average of 1s. |
| |
| Keywords: | |
|
| 点击此处可从《信号处理》浏览原始摘要信息 |
|
点击此处可从《信号处理》下载全文 |
|
