A VHA NLP Software Ecosystem for Collaborative Development and Integration

描述（由申请人提供）： 退伍军人事务部在电子病历方面投入了大量资金，并建立了一个收集所有患者医疗信息的全国性系统。目前，除了少数研究人员之外，所有人都无法访问医疗记录中的文本信息。为了从现有系统中获得最大价值，管理员和从业人员需要能够访问他们所需的文本信息。我们有责任从生物医学和患者护理资源中获得最大利益。临床自然语言处理（NLP）是解决方案的重要组成部分。 NLP的价值在生物医学领域已经得到了认可。这方面的证据包括为以下专注于临床NLP的国家计划提供资金：整合生物学和床边（i2b2），健康信息学研究联盟（CHIR），VA信息学和计算基础设施（芬奇），战略健康IT高级研究项目（SHARP）和电子病历和基因组学（eMERGE）。一方面，这些努力证明了对NLP研究的需求。他们开发了新的NLP工具，创建了带注释的数据集，开发了通用数据模型，共享了语义标签，甚至试点了一个原型软件生态系统。另一方面，信息学界的普遍共识是，由于缺乏互操作性和协作，处理和利用文本数据仍然具有挑战性。除非加快临床NLP的研发步伐，否则我们无法满足生物医学和健康服务研究界日益增长的NLP需求。 虽然协同开发有希望推进NLP科学和加快NLP工具生产的步伐，但缺乏一个充满活力的协作环境，吸引大量的 临床NLP开发人员和研究人员。在VA CHIR和芬奇的努力中，我们创建了一个名为V3NLP的原型NLP生态系统，该生态系统支持异构工具的互操作性和集成到VA研究和运营计划中。然而，缺乏促进协作和关键用户群所需的环境。在拟议的项目中，我们将研究V3NLP生态系统的现有和潜在用户的需求，以提高其实用性和易用性，并促进合作。 的最终目标 NLP生态系统旨在为临床文本产生新的、更准确的NLP方法。 这就需要很好地理解各类临床文本的特点以及现有方法的优缺点。由于大多数临床NLP解决方案都是由单独的用例和注释集合驱动的，因此所产生的解决方案针对特定NLP任务和所分析的文本语料库的特征进行了优化。由于有许多任务和语料库，临床NLP解决方案往往难以重用，特别是不同的开发人员。为了解决这个问题，我们将研究一个非常大的和异构的VA文本记录的集合的特征，以理解和模拟VA临床笔记中的子语言。这种系统和全面的子语言分析将在拟议的生态系统中发挥关键作用。它将指导新的临床NLP方法的开发以及现有解决方案的定制。 我们的总体目标是加速临床NLP的研究和开发。具体目标如下：（1）收集和分析NLP开发人员，健康信息学研究人员和健康服务研究人员的需求，以告知合作NLP生态系统的设计，这将有助于开发更准确的方法。 (2)设计和实施临床NLP生态系统，促进协作，加速研究和采用准确和可推广的NLP方法。 (3)进行全面的子语言分析，以指导基于VA文本注释创建适应性强的NLP工具和方法，以支持跨多个临床领域的文本处理和信息提取。

DESCRIPTION (provided by applicant): The VA has invested hugely in electronic medical records and has achieved a nationwide system that collects medical information from all patients. Currently, the textual information in the medical records is inaccessible to all but a small number of researchers. In order to obtain the highest value from this existing system, administrators and practitioners need to be able to access the textual information they need. It is our responsibility to get the most benefit from thi resource for biomedical and patient care. Clinical natural language processing (NLP) is an important part the solution. The value of NLP has been recognized in the biomedical domain. Evidence of this includes funding for the following national initiatives focused on clinical NLP: Integrating Biology and the Bedside (i2b2), Consortium for Health Informatics Research (CHIR), VA Informatics and Computing Infrastructure (VINCI), Strategic Health IT Advanced Research Projects (SHARP), and electronic Medical Records & Genomics (eMERGE). On the one hand, these efforts testify to the demand for NLP research. They have produced new NLP tools, created annotated datasets, developed common data models, shared semantic labels, and even piloted a prototype software ecosystem. On the other hand, the general consensus in the informatics community is that processing and utilizing textual data remains challenging due to lack of interoperability and collaboration. Unless the pace of research and development is accelerated in clinical NLP, we cannot meet the increasing NLP demand originated from the biomedical and health services research community. Although synergistic development has the promise of advancing the science of NLP and accelerating the pace of NLP tool production, there lacks a vibrant collaborative environment attracting participation of a significant number of clinical NLP developers and researchers. Within the VA CHIR and VINCI efforts, we have created a prototype NLP ecosystem called V3NLP that supports the interoperability and integration of heterogeneous tools into VA research and operational initiatives. The environment needed to foster collaboration and a critical mass of users, however, is lacking. In the proposed project, we will study the needs of existing and potential users of the V3NLP ecosystem to increase its utility and ease of adoption and to facilitate collaboration. The ultimate goal of an NLP ecosystem is to produce new and more accurate NLP methods for clinical text. This requires a good understanding of the characteristics of various types of clinical text and the strengths and weakness of existing methods. Because most clinical NLP solutions have been driven by individual use cases and note collections, the resultant solutions are optimized for the characteristics of the specific NLP tasks and text corpora analyzed. Since there are numerous tasks and corpora, clinical NLP solutions tend to be difficult to re-use, especially by different developers. To remedy this, we will research characteristics of a very large and heterogeneous collection of VA text records to understand and model sublanguages in VA clinical notes. This systematic and comprehensive sublanguage analysis will play a critical role in the proposed ecosystem. It will guide the development of new clinical NLP methods as well as the customization of existing solutions. Our general goal is to accelerate clinical NLP research and development. The specific aims are as follows: (1) Collect and analyze the needs of NLP developers, health informatics researchers and health services researchers to inform the design of a collaborative NLP ecosystem that will facilitate development of more accurate methods. (2) Design and implement a clinical NLP ecosystem that fosters collaboration and accelerates research and adoption of accurate and generalizable NLP methods. (3) Conduct a comprehensive sublanguage analysis to guide the creation of adaptable NLP tools and methods based on VA text notes to support text processing and information extraction across multiple clinical domains.