基于改进青蒿素优化算法的乳腺癌图像分割

Breast Cancer Image Segmentation Based on Improved Artemisinin Optimization Algorithm

摘要: 乳腺癌是最普遍的女性癌症之一，精确的图像分割能够提高病灶定位、特征提取以及病症诊断的可靠性，有效提高乳腺癌患者的存活率。多阈值最大熵分割方法因原理清晰、实现简便及适应性较强，被广泛应用于医学图像处理领域。然而，传统的阈值选择方法随着阈值数量增加，计算复杂度显著上升，难以兼顾分割精度与计算效率。为此，本文提出了一种改进的青蒿素优化算法(ERAO)，引入精英引导差分变异机制强化精英个体的信息利用能力，提升收敛速度与求解精度；而重启策略能够有效缓解种群停滞现象。通过IEEE CEC 2017基准函数集的对比实验验证了所提出算法优越的优化性能。进一步将其与2D Kapur熵相结合，应用于乳腺癌图像多阈值分割任务，实验结果表明该方法能够获得更优的阈值组合与更高质量的分割效果。

Abstract: Breast cancer is one of the most prevalent cancers among women, and accurate image segmentation can enhance the reliability of lesion localization, feature extraction, and disease diagnosis. The multi-threshold maximum entropy segmentation method is widely used in the field of medical image processing due to its clear principle, simple implementation, and strong adaptability. However, traditional threshold selection methods face significant computational complexity as the number of thresholds increases, making it difficult to balance segmentation accuracy and computational efficiency. To address this, this paper proposes an improved Elitist Artemis Optimization Algorithm (ERAO), which introduces an elitist-guided differential mutation mechanism to enhance the information utilization ability of elite individuals, improving convergence speed and solution accuracy. The restart strategy effectively alleviates the population stagnation phenomenon. Comparative experiments on the IEEE CEC 2017 benchmark function set verify the superior optimization performance of the proposed algorithm. Furthermore, by combining it with 2D Kapur’s entropy, it is applied to the multi-threshold segmentation task of breast cancer images. Experimental results show that this method can obtain a better threshold combination and higher-quality segmentation results.