Abstract
A stochastic model of the flow cytometer measurement process was developed to assess the nature of the observed coefficient of variation (CV%) of the mean fluorescence intensity (MFI) from a population of labeled microspheres (beads). Several sources of variability were considered: the total number of labels on a bead, the path through the laser beam, the optical absorption cross-section, the quantum yield, the numerical aperture of the collection optics, and the photoelectron conversion efficiency of the photomultiplier (PMT) cathode. The variation in the number of labels on a bead had the largest effect on the CV% of the MFI of the bead population. The variation in the path of the bead through the laser beam was minimized using flat-top lasers. The variability in the average optical properties of the labels was of minor importance for beads with sufficiently large number of labels. The application of the bead results to the measured CV% of labeled B cells indicated that the measured CV% was a reliable measure of the variability of antibodies bound per cell. With some modifications, the model can be extended to multicolor flow cytometers and to the study of CV% from cells with low fluorescence signal.