摘要
MR脑肿瘤图像为临床提供了丰富的信息诊断和生物医学研究资料,通过算法实现MR脑肿瘤图像的自动准确分类对医学分析和解释至关重要。文章提出了一种全新的方法,以判断用户给定的MR脑肿瘤图像是否正常。首先,利用小波变换对图像进行特征提取,然后通过主成分分析(PCA)降低特征向量维数,得到新的MR脑肿瘤图像并将其提交至不同核的支持向量机(KSVM)以对比其分类结果和精确度,最终确立分类效果最佳的最优模型。该方法可在一定程度上为病情诊断提供参考,从而提高诊断精确率并促进相关治疗的开展,进而保障患者的生命健康。
MR brain tumor images provide rich information for clinical diagnosis and biomedical research data.The automatic and accurate classification of MR brain tumor images through algorithms is crucial for medical analysis and interpretation.The article proposes a novel method to determine whether a user̓s given MR brain tumor image is normal.Firstly,wavelet transform is used to extract features from the image,and then principal component analysis(PCA)is used to reduce the dimensionality of the feature vectors.A new MR brain tumor image is obtained and submitted to different Kernel Support Vector Machine(KSVM)with different kernels to compare its classification results and accuracy.Finally,the optimal model with the best classification performance is established.This method can provide reference for disease diagnosis to a certain extent,thereby improving diagnostic accuracy and promoting the development of related treatments,thereby ensuring the life and health of patients.
作者
杨志成
梁霄
YANG Zhicheng;LIANG Xiao(School of Mathematics and Statistics,Hubei University of Arts and Science,Xiangyang,Hubei 441053,China)
出处
《计算机应用文摘》
2024年第8期102-104,107,共4页
Chinese Journal of Computer Application
基金
湖北文理学院大学生创新训练项目(S202310519002)。
