• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>International Conference on Bioinformatics and Computational Biology >A Data-driven Biomarker Computational Model for Lung Disease Classification
【24h】

A Data-driven Biomarker Computational Model for Lung Disease Classification

机译：肺病分类的数据驱动生物标志物计算模型

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

We develop a data-driven computational model that reliably classifies individual patient into one of 7 non-overlapping lung disease clinical types within our dataset: healthy non-smokers, smokers diagnosed with and without chronic obstructive pulmonary disease (COPD), adenocarcinoma, squamous cell carcinoma, cystic fibrosis, and acute lung injury. Panels of 12 cytokine blood serum biomarker measurements precisely classify both known and unknown patients into one of these distinct clinical types. Our model classifies clinical types and patients directly from the conditional relationships of noisy, incomplete, and variable protein concentration measurements, including outliers. Biomarker concentration measurements induce discrete state variables through a binning algorithm that exposes the conditional relationships and dependencies among the concentration data. A unique application of an XOR operation on the state space extracts the patterns identifying the set of distinctive features for each clinical type. Our model builds a discrete topological structure from a baseline data set, and is developed using several novel schemes designed specifically for this analysis. The result is a multidimensional space representing a characteristic set of states within each clinical type population.
机译：我们开发数据驱动的计算模型，可将个体患者分类为我们数据集中的7种非重叠肺病临床类型中的7种中的一种：健康的非吸烟者，诊断患有慢性阻塞性肺病（COPD），腺癌，鳞状细胞的吸烟者癌，囊性纤维化和急性肺损伤。 12个细胞因子血液血清生物标志物测量的面板精确地将已知和未知患者分类为这些独特的临床类型之一。我们的模型将临床类型和患者直接从嘈杂，不完整和可变蛋白质浓度测量的条件关系中进行，包括异常值。生物标志物浓度测量通过融合算法诱导离散状态变量，该算法暴露浓度数据之间的条件关系和依赖性。 XOR操作在状态空间上的独特应用提取识别每个临床类型的独特特征集的模式。我们的模型从基线数据集中构建了离散的拓扑结构，并使用专门为该分析设计的多种新颖的方案进行开发。结果是表示每个临床型群体内的特征套的多维空间。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号