Fluorescent microscope imaging generates images distorted by optics and illumination. The problem of dark current, flat-field and background correction is about removing these distortions for quantitative image analyses. We developed a computational method for background correction of terabyte-sized microscopy images. The results of such correction are displayed in Figure 1.
Increasing the acquisition area of an experiment to cover all cells and colonies is important to fully analyze cellular behavior. The increased large-scale experimental coverage comes with several computational challenges that include a background correction model over a mosaic of hundreds of spatially overlapping FOVs taken over the course of several days, during which the background diminishes as cell colonies grow.
The background correction poses challenges due to the complex interactions of cells, media, fluorescent biomarker and imaging light, and also due to the computational demands of processing images of growing cell colonies that cover entire FOVs without leaving any background pixels. Unlike single cell imaging where background areas around cells provide accurate estimates of background intensity, pluripotent stem cells grow as colonies of cells that merge with neighboring colonies over time as the culture progresses, and background areas in a colony culture become sparse at later times.
This work concerns a large-scale background correction method that (1) minimizes the Root Mean Square (RMS) error remaining after image correction, (2) maximizes the Signal-to-Noise Ratio (SNR) reached after downsampling, and (3) has a fast execution time.
An open-source tool is made available for free download. This technique works on raw fluorescent images. It requires the foreground/background segmentation of each tile. For more detailed information please refer to the following help documentation. (download)
The paper: J.Chalfoun et al., "Background Intensity Correction for Terabyte-Sized Time-Lapse Images", Journal of Microscopy, 2014 has a complete description of the method and its application on our datasets. (download pdf)
J.Chalfoun et al., "Background Intensity Correction for Terabyte-Sized Time-Lapse Images", Journal of Microscopy, 2014 (download pdf)
Matlab MCR - This link will take you to the Matlab website and is required to run the Matlab executable if you do not have a Matlab license. It can be downloaded free of charge.