中文摘要：

提出一种基于灰度共生矩阵和BP神经网络脑部CT图像模式分类模型的构建方法.运用灰度共生矩阵对图像进行纹理特征提取,将能量、惯性矩、局部稳定性、熵四项特征参数作为BP神经网络的输入向量,随机抽取160幅正常、异常属性的脑部CT图像,进行网络训练、测试评估.实验结果显示：分类平均正确率达99%,可为脑部CT图像的正常、异常初步分类提供参考依据.

英文摘要：

A pattern classification model for brain CT images is proposed, which is based on gray level co-occurrence matrix and BP neural network. First, we use the gray level co-occurrence matrix to extract the texture fea- ture of the image. Next, use the energy, inertia moment, local stability, entropy of four parameters as input vector. Finally, use the BP neural network to carry out the training test. The results showed that the average correct rate of classification was 99%. This method can provide reference tbr the normal and abnormal preliminat~~ classification of brain CT images.