arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2103.13511cs.LGcs.AIcs.CV
收藏

解决医学领域扩展的灾难性遗忘

Addressing catastrophic forgetting for medical domain expansion

Athinoula A. Martinos Center for Biomedical Imaging(阿西努拉·A.马蒂诺斯生物医学成像中心);Massachusetts General Hospital(麻省总医院);Indian Institute of Technology Delhi(印度德里理工学院);Massachusetts Institute of Technology(麻省理工学院);Stanford University(斯坦福大学)
Sharut Gupta, Praveer Singh, Ken Chang, Liangqiong Qu, Mehak Aggarwal, Nishanth Arun, Ashwin Vaswani, Shruti Raghavan, Vibha Agarwal, Mishka Gidwani, Katharina … 展开作者

Sharut Gupta, Praveer Singh, Ken Chang, Liangqiong Qu, Mehak Aggarwal, Nishanth Arun, Ashwin Vaswani, Shruti Raghavan, Vibha Agarwal, Mishka Gidwani, Katharina Hoebel, Jay Patel, Charles Lu, Christopher P. Bridge, Daniel L. Rubin, Jayashree Kalpathy-Cramer

摘要:

在现实世界的医疗环境中部署深度学习模型时,模型脆性是一个关键问题。一个在一个机构表现很好的模型,在其他机构测试时,表现可能会大幅下降。虽然汇集来自多个机构的数据集和再培训可能提供一个简单的解决方案,但它通常是不可行的,并且可能会损害患者的隐私。另一种方法是在对原始机构进行培训后,对后续机构的模型进行微调。值得注意的是,这种方法降低了原始机构的模型性能,这种现象被称为灾难性遗忘。在本文中,我们开发了一种方法来解决灾难性的遗忘的基础上弹性权重合并结合调制的批量归一化统计两种情况下:第一,扩展域从一个成像系统的数据到另一个成像系统的,第二,扩展域从一个大的多机构数据集到另一个单一的机构数据集。我们表明,我们的方法优于其他几个国家的最先进的方法,并提供理论上的理由批量归一化调制的功效。本研究的结果通常适用于任何需要域扩展的临床深度学习模型的部署。

英文摘要:

Model brittleness is a key concern when deploying deep learning models in real-world medical settings. A model that has high performance at one institution may suffer a significant decline in performance when tested at other institutions. While pooling datasets from multiple institutions and retraining may provide a straightforward solution, it is often infeasible and may compromise patient privacy. An alternative approach is to fine-tune the model on subsequent institutions after training on the original institution. Notably, this approach degrades model performance at the original institution, a phenomenon known as catastrophic forgetting. In this paper, we develop an approach to address catastrophic forget-ting based on elastic weight consolidation combined with modulation of batch normalization statistics under two scenarios: first, for expanding the domain from one imaging system's data to another imaging system's, and second, for expanding the domain from a large multi-institutional dataset to another single institution dataset. We show that our approach outperforms several other state-of-the-art approaches and provide theoretical justification for the efficacy of batch normalization modulation. The results of this study are generally applicable to the deployment of any clinical deep learning model which requires domain expansion.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑