Optimising image capture for low-light widefield quantitative fluorescence microscopy

Zane Peterkovic,Avinash Upadhya,Christopher Perrella,Admir Bajraktarevic,Ramses Bautista Gonzalez,Megan Lim,Kylie R Dunning,Kishan Dholakia
2024-10-25
Abstract:Low-light optical imaging refers to the use of cameras to capture images with minimal photon flux. This area has broad application to diverse fields, including optical microscopy for biological studies. In such studies, it is important to reduce the intensity of illumination to reduce adverse effects such as photobleaching and phototoxicity that may perturb the biological system under study. The challenge when minimising illumination is to maintain image quality that reflects the underlying biology and can be used for quantitative measurements. An example is the optical redox ratio which is computed from autofluorescence intensity to measure metabolism. In all such cases, it is critical for researchers to optimise selection and application of scientific cameras to their microscopes, but few resources discuss performance in the low-light regime. In this tutorial, we address the challenges in optical fluorescence imaging at low-light levels for quantitative microscopy, with an emphasis on live biological samples. We analyse the performance of specialised low-light scientific cameras such as the EMCCD, qCMOS, and sCMOS, while considering the differences in platform architecture and the contribution of various sources of noise. The tutorial covers a detailed discussion of user-controllable parameters, as well as the application of post-processing algorithms for denoising. We illustrate these concepts using autofluorescence images of live mammalian embryos captured with a two-photon light sheet fluorescence microscope.
Quantitative Methods,Image and Video Processing
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the optimization of image acquisition in low - light wide - field quantitative fluorescence microscopy. Specifically, the research aims to reduce the illumination intensity to lessen the impact of photobleaching and phototoxicity on biological systems, while maintaining the image quality that can reflect the underlying biological characteristics and can be used for quantitative measurement. For example, the optical redox ratio is calculated from the autofluorescence intensity and is used to measure metabolic activity. In these cases, it is crucial for researchers to optimize the selection and application of scientific cameras to microscopes, but there are fewer resources discussing performance in low - light environments. Therefore, by analyzing the performance of different low - light - dedicated scientific cameras (such as EMCCD, qCMOS, and sCMOS), considering the differences in platform architectures and the contributions of various noise sources, this paper discusses in detail the user - controllable parameters and the application of denoising post - processing algorithms to help researchers optimize the performance of their wide - field imaging systems, especially for biological imaging at low - light levels.