韩国科学技术学院Young-Gyu Yoon团队取得一项新突破。他们发现无偏倚预测可实现电压成像数据的精确去噪。该项研究成果发表在2023年9月18日出版的《自然-方法学》上。
研究人员研发了SUPPORT（利用成像数据中的时空信息进行统计无偏预测），这是一种用于消除电压成像数据中泊松-高斯噪声的自检测学习方法。SUPPORT基于以下理论，即电压成像数据中的像素值高度依赖于其时空相邻像素，即使其时间相邻帧本身并不能为统计预测提供有用的信息。
具有时空盲点的卷积神经网络捕获并利用上述依赖性精确去噪电压成像数据，其中某个时间范围内动作电位的存在无法通过其他帧的信息推断出来。通过仿真和实验，研究人员表明SUPPORT能够对电压成像数据和其他类型的显微镜图像精确去噪，同时保留场景中的基本动态。
附：英文原文
Title: Statistically unbiased prediction enables accurate denoising of voltage imaging data
Author: Eom, Minho, Han, Seungjae, Park, Pojeong, Kim, Gyuri, Cho, Eun-Seo, Sim, Jueun, Lee, Kang-Han, Kim, Seonghoon, Tian, He, Bhm, Urs L., Lowet, Eric, Tseng, Hua-an, Choi, Jieun, Lucia, Stephani Edwina, Ryu, Seung Hyun, Rzsa, Mrton, Chang, Sunghoe, Kim, Pilhan, Han, Xue, Piatkevich, Kiryl D., Choi, Myunghwan, Kim, Cheol-Hee, Cohen, Adam E., Chang, Jae-Byum, Yoon, Young-Gyu
Issue&Volume: 2023-09-18
Abstract: Here we report SUPPORT (statistically unbiased prediction utilizing spatiotemporal information in imaging data), a self-supervised learning method for removing Poisson–Gaussian noise in voltage imaging data. SUPPORT is based on the insight that a pixel value in voltage imaging data is highly dependent on its spatiotemporal neighboring pixels, even when its temporally adjacent frames alone do not provide useful information for statistical prediction. Such dependency is captured and used by a convolutional neural network with a spatiotemporal blind spot to accurately denoise voltage imaging data in which the existence of the action potential in a time frame cannot be inferred by the information in other frames. Through simulations and experiments, we show that SUPPORT enables precise denoising of voltage imaging data and other types of microscopy image while preserving the underlying dynamics within the scene.
DOI: 10.1038/s41592-023-02005-8
Source: https://www.nature.com/articles/s41592-023-02005-8
Nature Methods：《自然—方法学》，创刊于2004年。隶属于施普林格·自然出版集团，最新IF：47.99
官方网址：https://www.nature.com/nmeth/
投稿链接：https://mts-nmeth.nature.com/cgi-bin/main.plex
本期文章：《自然—方法学》：Online/在线发表
