基于CCQPSO算法的医学图像分割研究

Medical image segmentation based on CCQPSO

文中提出了一种动态上下文协同量子行为的粒子群优化(CCQPSO)算法,该算法在上下文向量与多个测量样本中的每个样本之间依次进行协同,每次协同后上下文向量完成自身更新,进而提高了粒子搜索能力和优化性能。将CCQPSO算法应用于人体胃腔CT的医学图像分割实验中,将医学图像像素划分为多个类别的大津算法(OTSU)分割参数。通过在基准数据集上的综合实验测试结果,验证了CCQPSO算法显著优于其他三种方法,并加快了收敛速度,降低了最终误差。在医学图像分割实例中演示了该方法,数值结果表明,CCQPSO在医学图像分割方面优于其他方法。

In this paper,a particle swarm optimization(CCQPSO)algorithm for dynamic context cooperative quantum behavior is proposed,which coordinates the context vectors with each sample in several measurement samples in turn.After each collaboration,the context vectors update themselves,which improves the particle search ability and optimization performance.CCQPSO algorithm is applied to medical image segmentation experiment of human gastric cavity CT.Medical image pixels are divided into several categories of Otsu algorithm(OTSU)segmentation parameters.The experimental results on the benchmark data set show that the CCQPSO algorithm is superior to the other three methods,and accelerates the convergence speed and reduces the final error.The method is demonstrated in an example of medical image segmentation.The numerical results show that CCQPSO is superior to other methods in medical image segmentation.