从以数据为中心的视角重新思考生物成像中的深度学习。
Rethinking deep learning in bioimaging through a data centric lens.
作者信息
Cao Jiajun, Wenzel Jan, Zhang Shanghang, Lampe Josephine, Wang Hongxiao, Yao Jiachen, Zhang Zhicheng, Zhao Shuo, Zhou Yu, Chen Chao, Schwaninger Markus, Yang Jufeng, Chen Danny Z, Chen Jianxu
机构信息
National Key Laboratory for Multimedia Information Processing, School of Computer Science, Peking University, Beijing, China.
Institute for Experimental and Clinical Pharmacology and Toxicology, Center of Brain, Behavior and Metabolism (CBBM), University of Lübeck, Lübeck, Germany.
出版信息
Npj Imaging. 2025 Jun 26;3(1):29. doi: 10.1038/s44303-025-00092-0.
Deep learning has become essential in bioimaging for tasks. By examining data-centric strategies in general AI and revisiting existing deep learning methods in bioimaging, we describe a prototypical “BioData-Centric AI” framework. For AI users in bioimaging, this framework promotes a more practical approach beyond simply annotating large datasets or relying on a universal model. For method developers, it highlights key research directions to enhance AI toolboxes for the bioimaging community.
深度学习在生物成像任务中已变得至关重要。通过研究通用人工智能中以数据为中心的策略并重新审视生物成像中现有的深度学习方法，我们描述了一个典型的“以生物数据为中心的人工智能”框架。对于生物成像领域的人工智能用户而言，该框架推动了一种超越简单注释大型数据集或依赖通用模型的更实用方法。对于方法开发者来说，它突出了增强生物成像社区人工智能工具箱的关键研究方向。
Zotero 插件