Biomedical Informatics

项目总结 生物医学信息学共享资源(BI-SR)的目标是提供协作和咨询 为斯坦福癌症研究所(SCI)成员提供服务，从而在癌症研究的所有阶段增加价值。 提供咨询，以确定临床数据，包括回顾和预期的，以及相关的 为支持以发现为导向的SCI的基础和翻译研究工作所需的基因组或试验数据 研究社区。BI-SR开发了用于分析和编码数据的检索和编码工具 注重数据质量和重现性的可视化。BI-SR是从以前的 生物统计和研究信息学SR于2017年推出，目前在MD Daniel Rubin的指导下运营。 BI-SR支持的核心资源是斯坦福癌症研究所研究数据库(SCIRDB)，其中包括 BI-SR员工为数据查询和数据自动售卖提供的完整服务菜单。SCIRDB数据库具有 继续开发以满足SCI成员的需求，以收集、整合和分析已确定的 并从电子健康记录(EPIC)和广泛的其他数据中分离出患者数据 资料来源包括病理学、分子诊断学、临床试验和其他临床相关数据库。这个 BI-SR还提供咨询服务，以确保其最佳使用。本供资期间的重大捐款 包括采用新数据(纳入成像和遗传数据)和开发新的 用于人工智能(AI)应用程序的联合学习以及数据改进的软件工具 通过开始向云迁移，为基础设施提供科学依据。所有计划的成员都使用BI-SR， 使用最多的是癌症治疗和血液系统恶性肿瘤项目。BI-SR已经 参与了250多个项目咨询请求，并向100多个癌症研究提供数据，导致 自2017年以来，超过15本以癌症为重点的出版物。其中两份是影响力很大的出版物。年度预算 的BI-SR是746,394美元，而CCSG的请求是55,879美元。因此，BI-SR利用广泛的 机构支持，只从CCSG基金中寻求7%的支持。BI-SR侧重于四个具体目标：(1) 维护并继续开发SCI研究数据库(SCIRDB)，以促进癌症研究；(2) 为SCI成员提供数据科学服务和技术，使癌症研究成为可能，(3)开发 基础设施和数据驱动的应用程序，以支持癌症群体和研究人员的需求，以及 (4)教育SCI成员了解BI-SR的技术为实现数据驱动提供的机会 癌症的发现和护理。

PROJECT SUMMARY The objective of the Biomedical Informatics Shared Resource (BI-SR) is to provide collaborative and consultative services to Stanford Cancer Institute (SCI) members and thereby to add value in all phases of cancer research. Consultation is provided on identifying clinical data, both retrospective and prospective, along with associated omic or trial data as necessary to support the basic and translational research efforts of a discovery-oriented SCI research community. The BI-SR has developed tools for the retrieval and coding of data for analysis and visualization with attention to data quality and reproducibility. The BI-SR was split out from the previous Biostatistics and Research Informatics SR in 2017 and now operates under the direction of Daniel Rubin, MD. The core resource supported by the BI-SR is the Stanford Cancer Institute Research Database (SCIRDB), with a full menu of services provided by the BI-SR staff for data query and data vending. The SCIRDB database has continued to be developed to meet the needs of SCI members to collect, integrate, and analyze both identified and de-identified patient data from the electronic health record (EPIC) and from a wide range of other data sources, including Pathology, Molecular Diagnostics, clinical trials, and other clinically-related databases. The BI-SR also provides consultative services to ensure its optimal use. Significant contributions in this funding period included the introduction of new data (incorporation of imaging and genetic data) and the development of new software tools for federated learning of artificial intelligence (AI) applications, as well as improvement in data science infrastructure by commencing migration to the cloud. Members of all programs use the BI-SR, with the heaviest use coming from the Cancer Therapeutics and Hematologic Malignancies programs. The BI-SR has engaged in over 250 project consultation requests and provided data to over 100 cancer studies, resulting in over 15 cancer-focused publications since 2017. Two of these were high-impact publications. The annual budget of the BI-SR is $746,394, yet the CCSG request is $55,879. Accordingly, the BI-SR leverages extensive institutional support and seeks only 7% from CCSG funds. The BI-SR is focused on four specific aims: (1) to maintain and continue to develop the SCI Research Database (SCIRDB) to catalyze cancer research, (2) to provide SCI members with data science services and technologies to enable cancer research, (3) to develop infrastructure and data-driven applications to support the needs of cancer disease groups and researchers, and (4) to educate SCI members about the opportunities BI-SR’s technologies provide for enabling data-driven cancer discovery and care.