Early Detection of Pancreatic Cancer with Human-in-the-Loop Deep Learning

通过人在环深度学习早期检测胰腺癌

• 批准号: 10592060
• 负责人: Jiheum Park
• 金额: $ 17.86万
• 依托单位: COLUMBIA UNIVERSITY HEALTH SCIENCES
• 依托单位国家: 美国
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-05-01 至 2028-04-30
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10592060
• 关键词: Address; Adoption; Algorithms; Architecture; Artificial Intelligence; Bayesian Network; Breast; Cancer Detection; Cancer Etiology; Categories; Cessation of life; Clinical; Collection; Colorectal; Data; Data Collection; Data Set; Development; Diagnosis; Early Diagnosis; Electronic Health Record; Future; Generations; Genetic Risk; Goals; Grouping; Health Benefit; Healthcare; Human; Individual; Institution; Intervention; Knowledge; Laboratories; Learning; Literature; Logistic Regressions; Lung; Malignant Neoplasms; Malignant neoplasm of pancreas; Medical; Medical Records; Medical center; Methodology; Methods; Modeling; Modernization; Morbidity - disease rate; Obesity; Output; Pancreatic Ductal Adenocarcinoma; Pathway interactions; Patients; Performance; Pharmaceutical Preparations; Population; Procedures; Process; Prostate; Protocols documentation; Public Health; Questionnaires; Reporting; Research; Risk; Risk Assessment; Risk Factors; Series; Signal Transduction; Signs and Symptoms; Structure; Surveys; System; Taxonomy; Techniques; Text; Time; Tobacco use; Training; Universities; Visualization; artificial neural network; cancer type; clinical practice; clinical predictors; deep learning; deep learning algorithm; deep learning model; deep neural network; health data; high risk; human-in-the-loop; improved; innovation; learning strategy; mortality; multimodal data; multimodality; neural network architecture; neurodevelopment; novel; novel strategies; pancreas imaging; patient variability; predictive modeling; prevent; risk prediction; screening; trend

The goal of this project is to leverage deep-learning algorithms on Electronic Health Records (EHRs) to improve the early detection of pancreatic ductal adenocarcinoma (PDAC), a malignancy with high mortality and morbidity. Compared to other major types of cancer in the U.S. (e.g., colorectal, prostate, breast, lung), PDAC has a uniquely high mortality with a 5-year survival of only 3%, largely due to the late stage of diagnosis and the aggressiveness of the malignancy. In this K25 application, we aim to develop novel structured methodologies for systematically incorporating human expert domain knowledge into the training procedure of deep-learning algorithms (“Human-in-the-Loop” approach) for improving the early detection of PDAC. The overarching hypothesis for this study, which has already been demonstrated in numerous other contexts, is that the “Human-in-the-Loop” approach imbues the deep learning in the PDAC prediction model with expert domain knowledge (e.g., clinical work-flows, statistical knowledge) to result in improved model performance as well as interpretability of results. The proposed research will accomplish three aims. In Aim 1, we will build preprocessing pipelines that are generalizable for analyzing multimodal data from different data collection systems (national, state, and institutional). The resultant pipelines will provide multimodal EHR deep embeddings optimized for deep learning applications in Aim 2 & 3. In Aim 2, we will investigate feature grouping strategies relying on information from clinical workflows and incorporate them into the deep learning prediction model. The proposed model will provide new clinical predictors represented by single or composite variables according to the grouping strategies. Examples in the literature show that such grouped predictors consistently have superior predictive power compared to their individual components. In Aim 3 we will study causal relationships between patient variables (including composite variables discovered in Aim 2) and the PDAC risk. We will use the Human-in-the-Loop approach where the possible causal relationships suggested from the deep learning model will be evaluated and corrected by human experts (e.g., clinicians, statisticians), to construct faithful Causal Bayesian Networks (CBNs) visualizing causal pathways from patient variables to PDAC risk. The resultant CBNs will be used as a framework for developing a risk assessment questionnaire to collect Patient- Generated Health Data (PGHD), which will be further evaluated and optimized in my future R01 focused on the development of a mobile survey application to efficiently collect PGHD and improve the early detection of PDAC. This proposal can potentially lead to new criteria for identifying high-risk patients for PDAC and inform targeted screening practices, that will likely be generalizable to other types of cancer.

该项目的目标是利用电子健康记录 (EHR) 的深度学习算法 提高胰腺导管腺癌 (PDAC) 的早期检测，这是一种死亡率很高的恶性肿瘤 与美国其他主要类型的癌症（例如结直肠癌、前列腺癌、乳腺癌、肺癌）相比， PDAC 具有独特的高死亡率，5 年生存率仅为 3%，这主要是由于诊断已处于晚期 以及恶性肿瘤的侵袭性。在此 K25 应用中，我们的目标是开发新颖的结构。 将人类专家领域知识系统地纳入培训程序的方法 用于改进 PDAC 早期检测的深度学习算法（“人在环”方法）。 这项研究的总体假设已经在许多其他情况下得到了证明， “人在环”方法将专家的深度学习融入到 PDAC 预测模型中 领域知识（例如，临床工作流程、统计知识），以提高模型性能： 以及结果的可解释性。 拟议的研究将实现三个目标：在目标 1 中，我们将构建预处理管道。 可推广用于分析来自不同数据收集系统（国家、州和州）的多模式数据 由此产生的管道将提供针对深度优化的多模式 EHR 深度嵌入。 目标 2 和 3 中的学习应用程序。在目标 2 中，我们将研究依赖于的特征分组策略 来自临床工作流程的信息并将其纳入深度学习预测模型中。 所提出的模型将根据以下内容提供由单个或复合变量表示的新临床预测因子 文献中的分组策略表明，这种分组预测变量具有一致的效果。 与它们的各个组成部分相比，具有更高的预测能力。在目标 3 中，我们将研究因果关系。 我们将使用患者变量（包括目标 2 中发现的复合变量）和 PDAC 风险之间的关系。 人在环方法，其中从深度学习中得出可能的因果关系 模型将由人类专家（例如人群、统计学家）评估和纠正，以构建忠实的模型 因果贝叶斯网络 (CBN) 可视化从患者变量到 PDAC 风险的因果路径。 由此产生的 CBN 将用作制定风险评估问卷的框架，以收集患者的信息 生成的健康数据（PGHD），将在我未来的 R01 中进一步评估和优化，重点关注 开发移动调查应用程序，以有效收集 PGHD 并提高早期发现率 PDAC。 该提案可能会制定新的标准来识别 PDAC 高危患者并告知 有针对性的筛查实践，可能会推广到其他类型的癌症。