Automatically determining cause of death from verbal autopsy narratives
Automatically determining cause of death from verbal autopsy narratives
复制标题
DOI：
10.1186/s12911-019-0841-9
复制
发表时间：
2019-07-09
影响因子：
3.5
通讯作者：
Hirst, Graeme
中科院分区：
文献类型：
--
作者：
Jeblee, Serena;Gomes, Mireille;Hirst, Graeme
背景：口头解剖（VA）是在医生未确定正式死因的情况下，对死亡前症状的事后书面采访报告。目前领先的自动化VA编码方法主要使用来自VA的结构化数据来分配CoD类别。我们提出了一种仅从VA自由文本叙述中自动确定CoD类别的方法。方法经过预处理和拼写校正后，我们的方法从叙述中提取词频计数，并将其作为四种不同机器学习分类器的输入：朴素贝叶斯、随机森林、支持向量机和神经网络。结果对于单个CoD分类，我们的最佳分类器的灵敏度为。15种化学需氧量的成人死亡率为770(目前报告的最佳敏感性为。57)。662个，有48个卫生组织类别。当在种群水平上预测CoD分布时，我们的最佳分类器实现。962病因特异性死亡率分数15个类别的准确性。908个类别，与领先的CoD分布估计方法相当。结论基于叙事的机器学习分类器在个体层面上的表现与基于结构化数据的分类器相当。此外，我们的方法表明，VA叙述提供了可以被机器学习系统用于自动CoD分类的重要信息。与基于结构化问卷的方法不同，该方法可以应用于任何死因推断数据集，而不考虑收集过程或原产国。
BackgroundA verbal autopsy (VA) is a post-hoc written interview report of the symptoms preceding a person's death in cases where no official cause of death (CoD) was determined by a physician. Current leading automated VA coding methods primarily use structured data from VAs to assign a CoD category. We present a method to automatically determine CoD categories from VA free-text narratives alone.MethodsAfter preprocessing and spelling correction, our method extracts word frequency counts from the narratives and uses them as input to four different machine learning classifiers: naive Bayes, random forest, support vector machines, and a neural network.ResultsFor individual CoD classification, our best classifier achieves a sensitivity of.770 for adult deaths for 15 CoD categories (as compared to the current best reported sensitivity of.57), and.662 with 48 WHO categories. When predicting the CoD distribution at the population level, our best classifier achieves.962 cause-specific mortality fraction accuracy for 15 categories and.908 for 48 categories, which is on par with leading CoD distribution estimation methods.ConclusionsOur narrative-based machine learning classifier performs as well as classifiers based on structured data at the individual level. Moreover, our method demonstrates that VA narratives provide important information that can be used by a machine learning system for automated CoD classification. Unlike the structured questionnaire-based methods, this method can be applied to any verbal autopsy dataset, regardless of the collection process or country of origin.