论文部分内容阅读
人体内微量元素浓度的变化预示人体健康状况的改变,对于肿瘤病人尤其重要。本工作收集了93份头发样品,其中包括48个前列腺癌病人和45个作为对照的正常人的头发样品。应用ICP-MS方法测量这些样品中20种微量元素组成,用主成分分析的统计模式识别方法(SPRA-PCA),分析测量结果,以寻求前列腺癌病人头发微量元素的变化特征。结果表明,在20种微量元素中,钙和磷的含量变化与前列腺癌密切相关。于是,用钙和磷的含量构建一个预报前列腺癌的可视化模型,可清晰辨别前列腺癌病人与正常人。为了验证模型的预报能力,用这个模型去预报一组新的样本,预报结果与临床诊断完全相同。
Changes in the concentration of trace elements in the human body indicates the change of human health is particularly important for cancer patients. This work collected 93 hair samples, which included hair samples from 48 prostate cancer patients and 45 healthy controls as controls. The trace elements in these samples were measured by ICP-MS method. Statistical pattern recognition (SPRA-PCA) was used to analyze the changes of hair trace elements in patients with prostate cancer. The results showed that the contents of calcium and phosphorus in 20 kinds of trace elements are closely related to prostate cancer. Thus, using calcium and phosphorus content to build a visual model for predicting prostate cancer, patients with prostate cancer can be clearly distinguished from normal. To validate the model’s predictive power, this model is used to predict a new set of samples with the same predictions as the clinical diagnosis.