Noise and colocalization in fluorescence microscopy: solving a problem

• The Pearson correlation coefficient (PCC) is regularly used in colocalization measurements, but it is sensitive to image noise. Images of fluorophores are usually degraded by Poisson and background noise and we have found that, even with apparently high quality images, the measured PCC is substantially understated, to the extent that the numbers become misleading. This means that ostensibly significant differences in the PCC between two populations may just reflect differing image quality while dissimilar levels of noise may mask significant differences. A new correction, based on measurements of image quality, derived from a pair of images for each fluorophore, aligns the measured PCC with the true PCC. Our method, the Replicate Based Noise Corrected Correlation (RBNCC), generates an accurate PCC even from poor images. It is highly photon efficient and therefore well suited for use in live cell fluorescence imaging or with rapidly bleaching fluorophores.

Publication and Content Type

vet (subject category)

for (subject category)