As you likely experienced, the results of coefficients calculations depend on both the method used to reduce noise (or as we call it to correct background) and the properties of the selected region of interest (ROI).
For example, the calculation results will differ for an image with background corrected using “Average” preset versus “Weak Fluorescence” preset. If the same background correction method was used, but the properties of selected ROI were different, for example in one case ROI was selected using Lasso tool and in another using Oval tool, then the results will differ again. The discrepancy of the results is usually minor to impact conclusions and can be remedied by re-using background correction settings and selecting the same types of ROI to ensure that the values of obtained coefficients between experiments are as comparable as possible. However, this issue may become a major obstacle in cases when colocalization changes together with the brightness of the images and the size and shape of the areas containing it.
To solve this problem, we have developed a new approach that corrects background automatically according to the selected ROI. We call it Smart Background Correction. This approach ties together background correction and ROI selection into a single step. Previously, we recommended using presets to correct background as a simplest and the most efficient way to do it. With presets, the background was corrected for the whole image. Then, users were selecting ROI and performing calculations of coefficients.
Different areas of the image, however, have varying biological properties and therefore different backgrounds. Correcting image background uniformly and then selecting ROI on the image is, therefore, not the ideal solution. With the new approach, the background is corrected specifically for the selected ROI considering its unique properties. Correction is performed differently depending on selected ROI, but using the same algorithm to ensure comparability of results. This represents a major improvement in reliability of colocalization analysis as well as simplifies its workflow.
A very smart algorithm within CoLocalizer will first set the initial threshold, then calculate the mean intensity values of image’s foreground versus its background, and, finally, figure out separate threshold values for different channels optimal for a particular selected ROI.
Our tests indicate that Smart Background Correction is vastly superior to presets, especially in cases when analyzing the changes of colocalization in dynamics. We therefore strongly recommend using this new approach for all your calculations.
Copyright © 2003-2018 CoLocalization Research Software.