Abstract:

Aiming at the high dimensional problem of pathological image morphometric features and the small amount of applied samples in medical field, this paper presents the ReliefF-HEPSO head and neck cancer pathological image feature selection algorithm. The algorithm constructs a framework of multi-level dimensionality reduction. Firstly, according to the correlation of features and categories, the ReliefF algorithm is used to determine different feature weights to achieve initial dimensionality reduction. Secondly, the evolutionary neural strategy (ENS) is used to enrich the particle populations diversity of the binary particle swarm optimization (BPSO). A hybrid binary evolutionary particle swarm optimization (HEPSO) algorithm is proposed to automatically search for the best feature    subsets of candidate feature subsets. Compared with 7 feature selection algorithms, the experiment proves that the   algorithm can effectively screen out the morphological features of high correlation pathology images, achieve rapid dimensionality reduction, and obtain higher classification performance with fewer features.

Key words: image features, small sample high dimension, ReliefF algorithm, evolutionary neural strategy (ENS), particle swarm optimization (PSO)

摘要：

针对原始病理图像经软件提取形态学特征后存在高维度，以及医学领域上样本的少量性问题，提出ReliefF-HEPSO头颈癌病理图像特征选择算法。该算法构建了多层次降维框架，首先根据特征和类别的相关性，利用ReliefF算法确定不同的特征权重，实现初步降维。其次利用进化神经策略（ENS）丰富二进制粒子群算法（BPSO）的种群的多样性，提出混合二进制进化粒子群算法（HEPSO）对候选特征子集完成最佳特征子集的自动寻找。与7种特征选择算法的实验对比结果证明，该算法能更有效筛选出高相关性的病理图像形态学特征，实现快速降维，以较少特征获得较高分类性能。

关键词: 图像特征, 小样本高维, ReliefF算法, 进化神经策略（ENS）, 粒子群算法（PSO）