Finding Medical Knowledge from Clinical Data based on Interpretations

本研究的重点是为医学专家获得更有趣的挖掘结果，在寻找血液检测与治疗效果之间的关系时，不仅要从目标临床数据本身进行数据挖掘，而且要将医学文献中的背景知识提取与专家的解释相结合。此外，根据生物医学对序列分析的高度关注，开发了针对病毒亚基因型和干扰素/利巴韦林治疗的半监督学习方法，以显示计算方法的潜力。

The focus of this research is to obtain more interesting mining results for the medical experts, so that not only the data mining from the target clinical data itself but also the combining the background knowledge extraction from medical litarature and experts’interpretations have been pursued in finding the relationships among blood tests and the treatment effects. dditionally, in accordance with the high attention to sequence analysis in biomedicine, semi-supervised learning methods specially for the virus subgenotypes and the interferon/ribavirin treatments were developed to show the potentials of computational approaches.