摘 要: 运用MATLAB软件对乳腺癌X光片做预处理、特征提取、基于SVM算法的分类器识别及辅助诊断。经反 复实验完成纹理特征提取、核函数选取、参数优化、样本优化等步骤,为影像学医师的诊断预测提供一种便捷有效的辅 助方式,减轻了医生的工作量,提高了医生的工作效率,降低漏诊误诊的机率。 |
关键词: 乳腺癌X光片;辅助诊断;SVM算法;特征提取 |
中图分类号: TP311
文献标识码: A
|
基金项目: 国家自然科学基金资助项目,项目编号:81560304;海南省自然科学基金资助项目,项目编号:20156224. |
|
Design and Implementation of the Breast X-Ray Assisted Diagnosis System Based on SVM Algorithm |
NIU Lin,ZHANG Yuwei,ZHANG Luxin
|
( Hainan Medical College, Haikou 571199, China)
|
Abstract: MATLAB software is used to preprocess breast cancer X-rays,feature extraction,classifier recognition based on SVM algorithm,assisted diagnosis and process information storage.After repeated experiments,texture feature extraction,kernel function selection,parameter optimization,sample optimization and other steps have been completed.It provides a convenient and effective assisting way for imaging doctors to diagnose and predict,reduces the workload and improves the efficiency of doctors,and reduces the chance of misdiagnosis. |
Keywords: breast X-ray;assisted diagnosis;SVM algorithm;feature extraction |