Natural Language Processing Platform for Cancer Surveillance

修改后的项目摘要/摘要部分 这项题为“癌症监测的自然语言处理平台”的提案是对PAR 16-349NLP(https://grants.nih.gov/grants/guide/pa-files/par-16-349.html)研究领域1的回应，该领域专门涉及自然语言处理工具的开发，以促进从与癌症登记活动有关的各种非结构化电子医疗记录中自动/无监督/最小监督地提取特定的离散癌症相关数据。它是通过多PI机制提交的--波士顿儿童医院/哈佛医学院的Guergana Svova教授、范德比尔特大学医学中心的Jeremy Warner博士、匹兹堡大学的Harry Hochheiser教授和肯塔基州癌症登记处/肯塔基大学的Eric Durbin教授。目前的提案建立在由美国国家癌症研究所癌症研究信息学工具(ITCR)计划(https://itcr.cancer.gov/)资助的先前工作的基础上。我们设想在我们迄今工作的基础上，推进临床表型数据的信息提取方法，以推动一种新的癌症监测范例，这将使基于医院、基于州和国家的癌症登记受益。在这种新的模式下，监测项目将使用这些方法来提高癌症报告的速度、准确性和简便性。拟议的DeepPhe*CR平台可以部署在本地站点或集中部署，并最终可以根据癌症监测界的需要集成到现有或新的可视化和抽象工具中。尽管之前已经有一些关于从各种数据流中自动提取表型的工作，包括特定类型癌症的临床叙述或用于癌症监测的个体变量，但拟议的工作将是朝着可推广的信息提取迈出的一步。这种通用性实现了可扩展性和可伸缩性。通过拟议项目的建模部分加强了互操作性，该部分以生物医学本体论、术语、社区通过的惯例和标准方面的最新进展为基础。我们计划与三个SEER癌症登记处合作，为我们的决策过程提供了大规模癌症监测的坚实基础。

Modified Project Summary/Abstract Section This UG3/UH3 proposal titled “Natural Language Processing Platform for Cancer Surveillance” is in response to Research Area 1 of PAR 16-349 (https://grants.nih.gov/grants/guide/pa-files/par-16-349.html) specifically addressing the development of natural language processing (NLP) tools to facilitate automatic/unsupervised/minimally supervised extraction of specific discrete cancer-related data from various types of unstructured electronic medical records (EMRs) related to the activities of cancer registries. It is submitted through a multi-PI mechanism – Prof. Guergana Savova from Boston Children’s Hospital/Harvard Medical School, Dr. Jeremy Warner from Vanderbilt University Medical Center, Prof. Harry Hochheiser from the University of Pittsburgh, and Prof. Eric Durbin from the Kentucky Cancer Registry/University of Kentucky. The current proposal builds on prior work funded by the NCI Informatics Tools for Cancer Research (ITCR) program (https://itcr.cancer.gov/ ). We envision building on our work to date to advance methods for information extraction of clinical phenotyping data needed to fuel a new cancer surveillance paradigm that would benefit hospital-based, state-based, and national cancer registries. In this new paradigm, surveillance programs would use the methods to enhance the speed, accuracy, and ease of cancer reporting. The proposed DeepPhe*CR platform could be deployed at local sites or centrally, and could eventually be integrated into existing or new visualization and abstraction tools as needed by the cancer surveillance community. Although there has been some previous work on automatic phenotype extraction from the various streams of data including the clinical narrative for specific types of cancer or individual variables for cancer surveillance, the proposed work will be a step towards a generalizable information extraction. This generalizability enables extensibility and scalability. Interoperability is reinforced through the modeling part of the proposed project which is grounded in most recent advances in biomedical ontologies, terminologies, community-adopted conventions and standards. Our planned partnership with three SEER cancer registries provides our decision-making processes with a solid foundation in large-scale cancer surveillance.