论文部分内容阅读
为了提高诊断正确率,大量医学研究集中于探讨肿瘤标志物正常值与非正常值的合适划分,即确定分割点。为此,该文提出了一种肿瘤标志物多分割点设置算法。通过正交实验和缩小搜索空间的方法,进一步提高了诊断正确率和优化收敛速度。在结直肠癌实际诊断数据中的应用效果表明,该算法能有效地提高诊断正确率,其诊断结果优于医学上的单分割点的诊断结果。
In order to improve the diagnostic accuracy, a large number of medical studies have focused on the appropriate division of the normal and abnormal values of tumor markers, ie to determine the segmentation points. For this reason, a multi-segment point setting algorithm for tumor markers is proposed in this paper. Through the orthogonal experiment and the method of reducing the search space, the diagnostic accuracy rate and the optimization convergence speed are further improved. The application results in the actual diagnostic data of colorectal cancer show that the algorithm can effectively improve the diagnostic accuracy, and its diagnostic result is better than the medical single diagnosis results.