基于电阻抗频谱的乳腺组织分类

Breast tissue classification based on electrical impedance frequency spectrum

针对乳腺癌早期辅助检测与诊断问题,提出一种改进的基于电阻抗频谱特性的乳腺组织分类算法。根据人体电阻抗特征数据特点,利用主成分分析法降低数据维数和去除噪声,然后采用支持向量机对数据进行分类。由此结合主成分分析和支持向量机,得到改进的分类算法。实验结果表明,改进的分类算法可以有效地对乳腺组织进行分类,准确率达80%以上,特别对乳腺癌和脂肪组织的分类准确率超高95%。

A method about the breast tissue classification is proposed based on the electrical impedance spectral characteristics with an aim for the early diagnosis problem of the breast disease. According to the characteristics of body electrical impedance data with the selection of the princi component analysis to reduce data dimensionality and noise removal, the breast tissue can pal be classified by the support vector machine classifier. Experiment results show that this method can classify the breast tissue effectively with an accuracy of more than 80%. Particularly, the classification efficiency for breast cancer and fatty tissue are more than 95 %.