OHSU Knight Federated Learning Network Prototype to Support Multimodal Cancer Models

支持多模式癌症模型的 OHSU Knight 联合学习网络原型

• 批准号: 10888579
• 负责人: BRIAN J DRUKER
• 金额: $ 29.98万
• 依托单位: OREGON HEALTH & SCIENCE UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 1997
• 资助国家: 美国
• 起止时间: 1997-08-01 至 2027-06-30
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10888579
• 关键词: Academic Medical Centers; Cancer Center; Cancer Model; Clinical; Collaborations; Computational Biology; Country; Data; Data Analyses; Development; Discrimination; Ethics; Evaluation; Future; Guidelines; Machine Learning; Malignant Neoplasms; Malignant neoplasm of prostate; Medical center; Modeling; Mutation; Patients; Privacy; Records; Research; Risk; Security; Site; Testing; Work; data de-identification; data quality; data sharing; deep learning model; federated learning; high dimensionality; learning network; multimodality; precision oncology; prototype; social stigma; tumor

PROJECT SUMMARY/ABSTRACT Traditional data-sharing models typically centralize de-identified data for analysis. Patient records have become more multifaceted and high-dimensional such that de-identification alone is inadequate to safeguard privacy. Privacy violations can result in harms to patients that can be both deontological (ethically problematic) or consequentialist (harm such as discrimination, financial consequences, and stigma). Current regulatory guidelines are not designed for the complexity of AI/ML. Moreover, each academic medical center will have their own ethics boards, governance models, privacy standards and security requirements. Risk of a privacy breach limits access and sharing of patient data — a barrier to research across cancer centers and academic medical centers. Given its outstanding strengths in precision oncology, computational biology, machine learning, clinical curation, data governance and data quality and regional data sharing, OHSU Knight Cancer Institute is uniquely situated to assist NCI in the development of a federated learning network. The long-term objective is to develop the best practices and guidance to support large-scale federated learning across cancer centers to achieve optimal benefit through broad data sharing. In the near term, the focus of this project is to support the development and testing of a federated learning network prototype. To do this, the OHSU Knight Cancer Institute proposes the following aims. Aim 1 is to develop a multi-modal deep learning model to predict tumor mutational burden in prostate cancer. Aim 2 is to implement and evaluate our federated learning prototype focusing on privacy, compliance and bias reduction. This will also include assessment of the impact of adversarial attacks and how well different defense strategies mitigate this. Finally, Aim 3 is the evaluation of the federated model on the federated learning network in collaboration with partner sites (other cancer centers, NCI).

项目总结/摘要 传统的数据共享模型通常集中去识别数据进行分析。患者 记录已经变得更加多方面和高维度， 不足以保护隐私。侵犯隐私可能会对患者造成伤害， 无论是义务论的（道德上有问题的）还是后果论的（歧视等伤害， 经济后果和耻辱）。目前的监管指南不是为 AI/ML的复杂性此外，每个学术医疗中心将有自己的道德委员会， 治理模式、隐私标准和安全要求。隐私泄露风险 限制访问和共享患者数据-这是跨癌症中心研究的障碍， 学术医疗中心鉴于其在精确肿瘤学方面的突出优势， 生物学、机器学习、临床策展、数据治理和数据质量以及区域数据 分享，OHSU骑士癌症研究所是独特的位置，以协助NCI在发展一个 联邦学习网络长期目标是制定最佳做法， 指导支持跨癌症中心的大规模联合学习， 通过广泛的数据共享。在近期，该项目的重点是支持 开发和测试联合学习网络原型。为此，OHSU 奈特癌症研究所提出了以下目标。目标1是开发一个多模式的深度 学习模型来预测前列腺癌的肿瘤突变负荷。目标2是执行和 评估我们的联邦学习原型，重点关注隐私、合规性和减少偏见。 这还将包括评估对抗性攻击的影响以及 防御战略减轻了这一点。最后，目标3是在 联合学习网络与合作伙伴网站（其他癌症中心，NCI）合作。

项目成果

Functional genomic landscape of acute myeloid leukaemia.

• DOI: 10.1038/s41586-018-0623-z
• 发表时间: 2018-10
• 期刊: Nature
• 影响因子: 64.8
• 作者: Tyner JW;Tognon CE;Bottomly D;Wilmot B;Kurtz SE;Savage SL;Long N;Schultz AR;Traer E;Abel M;Agarwal A;Blucher A;Borate U;Bryant J;Burke R;Carlos A;Carpenter R;Carroll J;Chang BH;Coblentz C;d'Almeida A;Cook R;Danilov A;Dao KT;Degnin M;Devine D;Dibb J;Edwards DK 5th;Eide CA;English I;Glover J;Henson R;Ho H;Jemal A;Johnson K;Johnson R;Junio B;Kaempf A;Leonard J;Lin C;Liu SQ;Lo P;Loriaux MM;Luty S;Macey T;MacManiman J;Martinez J;Mori M;Nelson D;Nichols C;Peters J;Ramsdill J;Rofelty A;Schuff R;Searles R;Segerdell E;Smith RL;Spurgeon SE;Sweeney T;Thapa A;Visser C;Wagner J;Watanabe-Smith K;Werth K;Wolf J;White L;Yates A;Zhang H;Cogle CR;Collins RH;Connolly DC;Deininger MW;Drusbosky L;Hourigan CS;Jordan CT;Kropf P;Lin TL;Martinez ME;Medeiros BC;Pallapati RR;Pollyea DA;Swords RT;Watts JM;Weir SJ;Wiest DL;Winters RM;McWeeney SK;Druker BJ
• 通讯作者: Druker BJ

IFNγ-activated dermal lymphatic vessels inhibit cytotoxic T cells in melanoma and inflamed skin.

• DOI: 10.1084/jem.20180654
• 发表时间: 2018-12-03
• 期刊: The Journal of experimental medicine
• 作者: Lane RS;Femel J;Breazeale AP;Loo CP;Thibault G;Kaempf A;Mori M;Tsujikawa T;Chang YH;Lund AW
• 通讯作者: Lund AW

Clonal tracing of Sox9+ liver progenitors in mouse oval cell injury.

• DOI: 10.1002/hep.27084
• 发表时间: 2014-07
• 期刊: HEPATOLOGY
• 影响因子: 13.5
• 作者: Tarlow, Branden D.;Finegold, Milton J.;Grompe, Markus
• 通讯作者: Grompe, Markus

A molecular case report: functional assay of tyrosine kinase inhibitors in cells from a patient's primary renal cell carcinoma.

• DOI: 10.4161/cbt.22960
• 发表时间: 2013-02
• 期刊: Cancer biology & therapy
• 影响因子: 3.6
• 作者: Kulesz-Martin MF;Lagowski J;Olson S;Wortham A;West T;Thomas G;Ryan C;Tyner JW
• 通讯作者: Tyner JW

Src and STAT3 inhibitors synergize to promote tumor inhibition in renal cell carcinoma.

• DOI: 10.18632/oncotarget.5971
• 发表时间: 2015-12-29
• 期刊: Oncotarget
• 作者: Lue HW;Cole B;Rao SA;Podolak J;Van Gaest A;King C;Eide CA;Wilmot B;Xue C;Spellman PT;Heiser LM;Tyner JW;Thomas GV
• 通讯作者: Thomas GV