Intelligent Systems Program, University of Pittsburgh, Pittsburgh, PA, United States; School of Computing and Information, University of Pittsburgh, Pittsburgh, PA, United States.
School of Computing and Information, University of Pittsburgh, Pittsburgh, PA, United States.
Int J Med Inform. 2019 Sep;129:81-87. doi: 10.1016/j.ijmedinf.2019.05.021. Epub 2019 Jun 6.
Radiologic imaging of trauma patients often uncovers findings that are unrelated to the trauma. These are termed as incidental findings and identifying them in radiology examination reports is necessary for appropriate follow-up. We developed and evaluated an automated pipeline to identify incidental findings at sentence and section levels in radiology reports of trauma patients.
We created an annotated dataset of 4,181 reports and investigated automated feature representations including traditional word and clinical concept (such as SNOMED CT) representations, as well as word and concept embeddings. We evaluated these representations by using them with traditional classifiers such as logistic regression and with deep learning methods such as convolutional neural networks (CNNs).
The best performance was observed using word embeddings with CNNs with F scores of 0.66 and 0.52 at section and sentence levels respectively. The F score was statistically significantly higher for sections compared to sentences (Wilcoxon; Z < 0.001, p < 0.05). Compared to using words alone, the addition of SNOMED CT concepts did not improve performance. At the sentence level, the F score improved significantly from 0.46 to 0.52 when using pre-trained embeddings (Wilcoxon; Z < 0.001, p < 0.05).
The results show that the best performance was achieved by using embeddings with CNNs at both sentence and section levels. This provides evidence that such a pipeline is capable of accurately identifying incidental findings in radiology reports in an automated manner.
创伤患者的放射影像学检查常常会发现与创伤无关的结果。这些结果被称为偶然发现，在放射学检查报告中识别这些发现对于进行适当的随访是必要的。我们开发并评估了一种自动化管道，用于在创伤患者的放射学报告中识别句子和段落级别的偶然发现。
我们创建了一个包含 4181 份报告的标注数据集，并研究了自动特征表示，包括传统的单词和临床概念（如 SNOMED CT）表示，以及单词和概念嵌入。我们使用传统分类器（如逻辑回归）和深度学习方法（如卷积神经网络（CNN））来评估这些表示。
在句子和段落级别，使用单词嵌入和 CNN 的最佳性能分别为 0.66 和 0.52 的 F 分数。与句子相比，F 分数在段落上的表现显著更高（Wilcoxon；Z < 0.001，p < 0.05）。与仅使用单词相比，添加 SNOMED CT 概念并不能提高性能。在句子级别，当使用预训练的嵌入时，F 分数从 0.46 显著提高到 0.52（Wilcoxon；Z < 0.001，p < 0.05）。
结果表明，在句子和段落级别，使用 CNN 进行嵌入的性能最佳。这表明该管道能够以自动化方式准确识别放射学报告中的偶然发现。
