论文部分内容阅读
摘要目的乳腺癌的个体化治疗是基于组织学和免疫组织化学(IHC)亚型的。MRI纹理分析(TA)在非浸润性病灶亚型分类中可能有一定作用。材料与方法新诊断的原发性乳腺癌病人治疗前行动态对比增强乳腺MRI检查。TA是应用共生矩阵(COM)特征,通过回顾性训练集数据创建一个模型,然后前瞻性地应用于测试集。在未知乳腺病理的情况下
Abstract Objective The individualized treatment of breast cancer is based on histological and immunohistochemical (IHC) isoforms. MRI texture analysis (TA) may play a role in the classification of noninvasive lesion subtypes. MATERIALS AND METHODS Newly diagnosed patients with primary breast cancer underwent dynamic contrast-enhanced breast MRI before treatment. TA is a co-occurrence matrix (COM) feature that creates a model from retrospective training set data and then applies it prospectively to the test set. In the case of unknown breast pathology