Abstract: In order to improve the brain CT image classification accuracy, this paper proposes brain CT mage classification model（IHS-LSSVM） based on the least squares support vector machine and harmony search algorithm. Firstly, the LSSVM parameters are taken as different musical tone combination, and then the harmony search algorithm is used to find the optimal parameters, and the optimal position adjustment strategy is introduced to enhance the ability of jumping out of local minima, the brain CT image classification model is established according to the optimal parameters, and the performance of the model is tested. The simulation results show that, compared with the other models, IHS-LSSVM not only improves the image classification accuracy, but also accelerates the classification speed, so it is an effective brain CT image classification model.

Key words: medical image classification, least squares support vector machines, harmony search algorithm, particle swarm optimization algorithm

摘要： 为了提高脑CT图像的分类正确率，针对分类器中的最小二乘支持向量机（LSSVM）参数优化问题，提出一种改进和声搜索算法优化LSSVM的脑CT图像分类模型（IHS-LSSVM）。将LSSVM参数看作不同乐器的声调组合，通过和声搜索算法的“调音”找到最优参数，并在寻优过程中引入粒子群算法的最优位置更新策略，增强了算法跳出局部极小值的能力，根据最优参数建立脑CT图像分类模型，并对模型的性能进行仿真测试。仿真结果表明，相对于对比模型，IHS-LSSVM不仅提高了脑CT图像分类正确率，而且加快分类速度，是一种有效的脑CT图像分类模型。

关键词: 脑CT图像分类, 最小二乘支持向量机, 和声搜索算法, 粒子群优化算法