局部兴奋全局抑制神经元振荡网络在医学图像分割中的应用

视觉皮层神经元之间存在的局部兴奋全局抑制机制能够使神经元根据接受到的视觉信息产生相应的振荡活动。在深入分析和探讨基于此生理振荡活动的具有局部兴奋全局抑制机制的视觉皮层神经元振荡网络模型的基础之上，运用最小交叉熵原理提出了一种自动确定参数的图像分割新算法。该算法应用于医学图像分割试验中，试验结果表明，这种以生物视觉感知机理为基础的神经计算方法可有效的检测出不同目标。由于该振荡网络具有高速并行运算，便于硬件实现等特点，因此这种方法在图像实时处理中具有很大的潜力和应用前景。

The Application of Locally Excitatory Globally Inhibitory Neuronal Oscillator Network on Medical Image Segmentation

Through the mechanism called locally excitatory globally inhibitory existing among the visual cortical neurons, neurons can oscillate in terms of visual information. By deeply analyzing the model of locally excitatory globally inhibitory visual cortical neuronal oscillator network with the mechanism among visual neurons and the theory of cross entropy, an adaptive parameter setting method of image segmentation was built in this work. Experimental results showed that the method based on biological visual mechanisms could detect different targets effectively. The cellular neural networks are uniquely suited for high-speed parallel computation and easy to be implemented in hardware, thus the proposed model is expected to have great potential in real-time image processing.