研究动态
Articles below are published ahead of final publication in an issue. Please cite articles in the following format: authors, (year), title, journal, DOI.

接力学习:用于临床多站点深度学习的物理安全框架。

Relay learning: a physically secure framework for clinical multi-site deep learning.

发表日期:2023 Nov 04
作者: Zi-Hao Bo, Yuchen Guo, Jinhao Lyu, Hengrui Liang, Jianxing He, Shijie Deng, Feng Xu, Xin Lou, Qionghai Dai
来源: Brain Structure & Function

摘要:

大数据是跨各个领域构建现实世界深度学习系统的基石。在医学和医疗保健领域,单个临床站点缺乏足够的数据,因此需要多个站点的参与。不幸的是,对数据安全和隐私的担忧阻碍了跨站点的数据共享和重用。现有的多站点临床学习方法在很大程度上取决于网络防火墙和系统实施的安全性。为了解决这个问题,我们提出了中继学习,这是一种安全的深度学习框架,可以将临床数据与外部入侵者物理隔离,同时仍然利用多站点大数据的优势。我们展示了接力学习在不同疾病和解剖结构的三个医疗任务中的功效,包括视网膜眼底结构分割、纵隔肿瘤诊断和大脑中线定位。我们通过多站点验证和外部验证将接力学习的性能与替代解决方案进行比较来评估接力学习。合并来自 21 个医疗主机(包括 7 个外部主机)的总共 41,038 个分布不均匀的医学图像,我们观察到中继学习在所有三项任务中的性能显着提高。具体来说,它在视网膜眼底分割、纵隔肿瘤诊断和脑中线定位方面分别实现了 44.4%、24.2% 和 36.7% 的平均性能提升。值得注意的是,接力学习在外部测试集上的表现甚至优于集中学习。同时,Relay Learning在本地保留数据主权,无需跨站点网络连接。我们预计接力学习将彻底改变临床多站点协作并重塑未来的医疗保健格局。© 2023。作者。
Big data serves as the cornerstone for constructing real-world deep learning systems across various domains. In medicine and healthcare, a single clinical site lacks sufficient data, thus necessitating the involvement of multiple sites. Unfortunately, concerns regarding data security and privacy hinder the sharing and reuse of data across sites. Existing approaches to multi-site clinical learning heavily depend on the security of the network firewall and system implementation. To address this issue, we propose Relay Learning, a secure deep-learning framework that physically isolates clinical data from external intruders while still leveraging the benefits of multi-site big data. We demonstrate the efficacy of Relay Learning in three medical tasks of different diseases and anatomical structures, including structure segmentation of retina fundus, mediastinum tumors diagnosis, and brain midline localization. We evaluate Relay Learning by comparing its performance to alternative solutions through multi-site validation and external validation. Incorporating a total of 41,038 medical images from 21 medical hosts, including 7 external hosts, with non-uniform distributions, we observe significant performance improvements with Relay Learning across all three tasks. Specifically, it achieves an average performance increase of 44.4%, 24.2%, and 36.7% for retinal fundus segmentation, mediastinum tumor diagnosis, and brain midline localization, respectively. Remarkably, Relay Learning even outperforms central learning on external test sets. In the meanwhile, Relay Learning keeps data sovereignty locally without cross-site network connections. We anticipate that Relay Learning will revolutionize clinical multi-site collaboration and reshape the landscape of healthcare in the future.© 2023. The Author(s).