Regional Oncology Research Center (LLMs for Unstructured Data Extraction)

区域肿瘤学研究中心（非结构化数据提取法学硕士）

• 批准号: 10891024
• 负责人: WILLIAM George NELSON
• 金额: $ 30万
• 依托单位: JOHNS HOPKINS UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-06-01 至 2027-05-31
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10891024
• 关键词: Address; Artificial Intelligence; Clinical; Clinical Data; Clinical Management; Data Commons; Data Element; Diagnosis; Engineering; Equity; Healthcare; Individual; Information Retrieval; Language; Malignant Neoplasms; Manuals; Modeling; Molecular; Natural Language Processing; Oncology; Online Systems; Output; Pathologic; Pathology Report; Patients; Performance; Play; Process; Prognosis; Reporting; Research; Research Design; Role; Structure; Supervision; Techniques; Text; Training; Work; cancer diagnosis; cohort; compare effectiveness; design; heuristics; improved; innovation; interdisciplinary approach; medical schools; model building; model development; neoplasm registry; precision medicine; unstructured data

Abstract Artificial intelligence (AI) has the potential to revolutionize healthcare by leveraging clinical data to advance research and improve oncology practice. Within free-text pathology reports, crucial information about primary cancer diagnoses and evolving molecular features is embedded. Extracting and interpreting this information accurately is essential for determining cancer stage, which plays a decisive role in prognosis and guiding clinical management. Although natural language processing (NLP) techniques have been applied to extract focused information from pathology reports, there is still a need for adaptable, generalizable, and interpretable strategies to enhance clinical data abstraction. To address this need, we propose a multidisciplinary approach to develop an integrative clinical information extraction pipeline. This work aims to improve, assess, and enhance the abstraction of relevant features of pathological diagnosis from pathology reports by leveraging large language models. Our research design involves several steps. First, we will establish a diverse and equitable cohort of patients from our Cancer Registry and collect free-text pathology reports, along with structured clinical data obtained from the Johns Hopkins School of Medicine Precision Medicine Analytics Platform (PMAP) Data Commons. Next, we will employ an information extraction platform to identify pathological features from the reports. This platform will utilize a suite of models, including BERT-like models, GPT-3.5, and GPT-4, provided by Microsoft, specifically designed for identifying key cancer attributes. Subsequently, we will evaluate the output of individual models using the CASPER interactive model development framework, enhancing and refining the results through heuristics and weak supervision. The augmented model output will be presented through a web-based user interface, allowing expert curators to provide further input. We will then compare the effectiveness of each CASPER-augmented model and its derived pathological features against the established gold standard annotations from the Cancer Registry. Finally, we will enhance the GPT-based language models based on the assessment, curation, and comparison process, employing prompt engineering techniques to improve performance and mitigate bias.

抽象的 人工智能 (AI) 有潜力利用临床数据推动医疗保健领域的变革 研究和改进肿瘤学实践。在自由文本病理报告中，有关原发性的重要信息 癌症诊断和不断发展的分子特征已被嵌入。提取和解释此信息 准确地确定癌症分期至关重要，对预后和指导临床起着决定性作用 管理。尽管自然语言处理（NLP）技术已被应用于提取重点 来自病理报告的信息，仍然需要适应性强、可推广和可解释的策略 加强临床数据抽象。为了满足这一需求，我们提出了一种多学科方法来开发 综合临床信息提取管道。这项工作旨在改进、评估和增强 利用大语言从病理报告中提取病理诊断的相关特征 模型。 我们的研究设计涉及几个步骤。首先，我们将建立一个多样化且公平的患者群体 从我们的癌症登记处收集自由文本病理学报告，以及从以下来源获得的结构化临床数据 约翰霍普金斯大学医学院精准医学分析平台 (PMAP) 数据共享。接下来，我们 将采用信息提取平台从报告中识别病理特征。该平台将 利用 Microsoft 提供的一套模型，包括类似 BERT 的模型、GPT-3.5 和 GPT-4 旨在识别关键的癌症属性。随后，我们将评估各个模型的输出 使用CASPER交互式模型开发框架，通过以下方式增强和完善结果 启发式和弱监督。增强模型输出将通过基于网络的用户呈现 界面，允许专家策展人提供进一步的输入。然后我们将比较每种方法的有效性 CASPER 增强模型及其根据既定金标准得出的病理特征 来自癌症登记处的注释。最后，我们将在以下基础上增强基于 GPT 的语言模型： 评估、策划和比较过程，采用及时的工程技术来改进 性能并减少偏见。