Enhanced Metadata Design, Architecture, and Learning (MeDAL) for Development of Generalizable Deep Learning-based Predictive Analytics from Electronic Health Records

增强元数据设计、架构和学习 (MeDAL)，用于根据电子健康记录开发基于深度学习的通用预测分析

• 批准号: 10265157
• 负责人: SHAMIM NEMATI
• 金额: $ 39.4万
• 依托单位: UNIVERSITY OF CALIFORNIA, SAN DIEGO
• 依托单位国家: 美国
• 财政年份: 2020
• 资助国家: 美国
• 起止时间: 2020-09-25 至 2022-08-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10265157
• 关键词: Accident and Emergency department; Acute Renal Failure with Renal Papillary Necrosis; Adverse event; Algorithms; Antibiotics; Architecture; Area Under Curve; Artificial Intelligence; Biometry; Calibration; Caring; Cessation of life; Code; Collaborations; Collection; Computer software; Computers; Confidence Intervals; Critical Illness; Data; Data Element; Data Provenance; Data Set; Data Sources; Development; Devices; Effectiveness; Electronic Health Record; Ensure; Environment; Evaluation; Event; Expenditure; Fast Healthcare Interoperability Resources; Frequencies; Funding; General Hospitals; Goals; Health; Health Expenditures; Heterogeneity; Hospital Costs; Hospital Mortality; Hospitalization; Hospitals; Hour; Incidence; Individual; Infection; Inflammation; Institution; Intensive Care Units; Learning; Length of Stay; Life; Measurement; Measures; Medical; Metadata; Methodology; Methods; Modeling; Morbidity - disease rate; Organ failure; Outcome; Patient-Focused Outcomes; Patients; Pattern; Performance; Pharmaceutical Preparations; Pilot Projects; Population; Predictive Analytics; Prevention; Process; Reproducibility; Research; Research Personnel; Risk; Risk Assessment; Risk Estimate; Savings; Sepsis; Septic Shock; Site; Source; Standardization; Testing; Time; Training; Uncertainty; Variant; Work; acute care; aging population; authority; base; cloud based; deep learning; demographics; design; improved; interest; novel; patient population; patient response; personalized care; portability; prediction algorithm; predictive modeling; pressure; prevent; prospective; research and development; response; septic patients; theories; tool; treatment optimization; ward

Sepsis, Septic Shock, and Acute Kidney Injury (AKI) are among the top causes of hospital mortality, morbidity, and an increase in duration and cost of hospitalization. Successful prevention and management of these conditions rely on the ability of clinicians to estimate the risk, and ideally, to anticipate and prevent these events. Acute care settings and in particular intensive care units (ICUs) provide an environment where an immense amount of data is acquired, and it is expected that with the advent of wearables and biometric patches even more data will be available in such settings. But at present, very little of these data are used effectively to prognosticate, and the existing predictive analytics risk scores suffer from lack of generalizability across institutions and performance degradation within the same institution over time. The PIs on this proposal recently demonstrated that a Deep Learning-based algorithm can reliably predict new sepsis cases in the emergency departments, general hospital wards, and ICUs by as much as 4-6 hours in advance and an area under the curve (ROC) of 0.85-0.90. Furthermore, through a 2-year pilot study funded via Biomedical Advanced Research and Development Authority (BARDA), we recently joined forces in a multicenter academic consortium to retrospectively validate this algorithm at each site. Our collaboration has resulted in a multi-center longitudinal EHR dataset of critically ill patients and has generated several important questions and findings related to design of portable and generalizable predictive analytics algorithms that are robust to problems arising from gaps, errors, and biases in electronic health records (EHRs) due to workflow-related factors (e.g. staffing-level), and heterogeneity of patient populations and measurement devices. We propose to continue our prior work by designing new deep learning architectures that are more robust to data missingness and biases introduced through the variability in process of care, 2) development of new learning methodologies to improve generalizability of the proposed models under data/population drifts (aka distributional changes), 3) enhanced metadata design to assist in quantifying ‘conditions for use’ of such algorithms via algorithmic controls, and 4) HL7 and FHIR-based prospective implementation and testing of these methodologies to provide real-world evidence for the effectiveness of the proposed approaches. Ultimately, these novel methodologies and tools will enhance our ability to use EHR and other types of continuously measured longitudinal data to predict adverse events, assess patients’ response to therapy, and optimize and personalize care at the beside.

败血症，败血性休克和急性肾脏损伤（AKI）是医院死亡率的主要原因之一， 发病率，以及住院持续时间和成本的增加。成功预防和管理 在这些情况下 这些事件。急性护理环境，尤其是重症监护病房（ICU）提供了一个环境 在获取大量数据的地方，可以预期随着可穿戴设备的冒险 生物识别补丁甚至可以在此类设置中获得更多数据。但是目前，这些数据很少 有效地用于证明烯酸盐，现有的预测分析风险评分缺乏 随着时间的流逝，机构之间的概括性和绩效下降。 该提案的PI最近表明，基于深度学习的算法可以可靠地预测 急诊科，综合医院病房和ICU的新败血症病例高达4-6 提前几小时，曲线（ROC）下的面积为0.85-0.90。此外，通过2年的飞行员 通过生物医学高级研发局（Barda）资助的研究，我们最近加入了 多中心学术联盟中的力量回顾性地验证了每个地点的该算法。我们的 协作导致了一个多中心的纵向EHR数据集的重症患者，并且有 产生了几个重要的问题和发现，与便携式和可推广的设计有关 预测分析算法对因差距，错误和偏见引起的问题而强大的算法 由于与工作流有关的因素（例如人员配置级）和异质性，电子健康记录（EHRS） 患者人群和测量设备。 我们建议通过设计新的深度学习体系结构来继续我们的先前工作 通过护理过程的变异性引入的数据丢失和偏见，2） 在数据/人群下提高拟议模型的普遍性的新学习方法 戏剧（又称分布变化），3）增强的元数据设计以帮助量化“使用条件” 通过算法控制和4）HL7和基于FHIR的前瞻性实施和 测试这些方法以提供现实世界的证据证明所提出的有效性 方法。最终，这些新颖的方法和工具将增强我们使用EHR和其他其他的能力 连续测量的纵向数据的类型以预测不良事件，评估患者对 治疗，并在旁边进行优化和个性化护理。