Broder Josef, Majumder Anirban, Porter Erika, Srinivasamoorthy Ganesh, Keith Charles, Lauderdale James, Sornborger Andrew
Center for Applied Mathematics, Cornell University, Ithaca, NY 14853, USA.
J Opt Soc Am A Opt Image Sci Vis. 2007 Sep;24(9):2921-31. doi: 10.1364/josaa.24.002921.
Ratiometric fluorescent indicators are becoming increasingly prevalent in many areas of biology. They are used for making quantitative measurements of intracellular free calcium both in vitro and in vivo, as well as measuring membrane potentials, pH, and other important physiological variables of interest to researchers in many subfields. Often, functional changes in the fluorescent yield of ratiometric indicators are small, and the signal-to-noise ratio (SNR) is of order unity or less. In particular, variability in the denominator of the ratio can lead to very poor ratio estimates. We present a statistical optimization method for objectively detecting and estimating ratiometric signals in dual-wavelength measurements of fluorescent, ratiometric indicators that improves on standard methods. With the use of an appropriate statistical model for ratiometric signals and by taking the pixel-pixel covariance of an imaging dataset into account, we are able to extract user-independent spatiotemporal information that retains high resolution in both space and time.
比率荧光指示剂在生物学的许多领域正变得越来越普遍。它们用于在体外和体内对细胞内游离钙进行定量测量，以及测量膜电位、pH值和许多子领域研究人员感兴趣的其他重要生理变量。通常，比率指示剂荧光产量的功能变化很小，信噪比（SNR）为单位1或更小。特别是，比率分母中的变异性可能导致非常差的比率估计。我们提出了一种统计优化方法，用于在荧光比率指示剂的双波长测量中客观地检测和估计比率信号，该方法优于标准方法。通过使用适用于比率信号的统计模型并考虑成像数据集的像素间协方差，我们能够提取在空间和时间上都保持高分辨率的与用户无关的时空信息。
