Machine-learned solutions for three stages of clinical information extraction: the state of the art at i2b2 2010.
Machine-learned solutions for three stages of clinical information extraction: the state of the art at i2b2 2010.
复制标题
DOI：
10.1136/amiajnl-2011-000150
复制
发表时间：
2011-09
期刊：
影响因子：
--
通讯作者：
Zhu X
中科院分区：
文献类型：
--
作者：
de Bruijn B;Cherry C;Kiritchenko S;Martin J;Zhu X
随着临床文本挖掘技术的不断成熟，其作为患者护理和临床研究创新技术的潜力正在成为现实。该过程的一个关键部分是对现实临床叙事的自然语言处理方法进行严格的基准测试。在本文中，作者描述了加拿大国家研究委员会在2010 i2b2挑战评估中的三个最先进的文本挖掘应用程序的设计和性能。这三个系统完成了临床信息提取的三个关键步骤：(1)从出院摘要和进度记录中提取医疗问题、检查和治疗；（二）医疗问题主张的分类；(3)医学概念之间关系的分类。机器学习系统使用大维度的特征包来执行这些任务，这些特征包来自文本本身和外部来源：UMLS、ctake和Medline。使用微平均f分数来衡量每个子任务的性能，这是通过比较系统注释和测试集上的真值注释来计算的。该系统在所有提交的系统中排名较高，f分如下：概念提取0.8523（排名第一）；断言检测0.9362（排名第一）；关系检测0.7313（排名第二）。对于所有任务，我们发现引入广泛的功能是成功的关键。重要的是，我们选择的机器学习算法允许我们在特征设计中具有通用性，并且可以在不过度拟合和不遇到计算资源瓶颈的情况下引入大量特征。
As clinical text mining continues to mature, its potential as an enabling technology for innovations in patient care and clinical research is becoming a reality. A critical part of that process is rigid benchmark testing of natural language processing methods on realistic clinical narrative. In this paper, the authors describe the design and performance of three state-of-the-art text-mining applications from the National Research Council of Canada on evaluations within the 2010 i2b2 challenge. The three systems perform three key steps in clinical information extraction: (1) extraction of medical problems, tests, and treatments, from discharge summaries and progress notes; (2) classification of assertions made on the medical problems; (3) classification of relations between medical concepts. Machine learning systems performed these tasks using large-dimensional bags of features, as derived from both the text itself and from external sources: UMLS, cTAKES, and Medline. Performance was measured per subtask, using micro-averaged F-scores, as calculated by comparing system annotations with ground-truth annotations on a test set. The systems ranked high among all submitted systems in the competition, with the following F-scores: concept extraction 0.8523 (ranked first); assertion detection 0.9362 (ranked first); relationship detection 0.7313 (ranked second). For all tasks, we found that the introduction of a wide range of features was crucial to success. Importantly, our choice of machine learning algorithms allowed us to be versatile in our feature design, and to introduce a large number of features without overfitting and without encountering computing-resource bottlenecks.
登录
查看更多内容
影响因子：
4.5
作者：
Harkema H;Dowling JN;Thornblade T;Chapman WW
通讯作者：
Chapman WW
DOI：
10.1145/1961189.1961199
发表时间：
2011-01-01
影响因子：
5
作者：
Chang, Chih-Chung;Lin, Chih-Jen
通讯作者：
Lin, Chih-Jen
DOI：
10.1136/amiajnl-2011-000203
发表时间：
2011-09-01
影响因子：
6.4
作者：
Uzuner, Oezlem;South, Brett R.;DuVall, Scott L.
通讯作者：
DuVall, Scott L.
DOI：
10.1136/jamia.2009.002733
发表时间：
2010-05-01
影响因子：
6.4
作者：
Aronson, Alan R.;Lang, Francois-Michel
通讯作者：
Lang, Francois-Michel