Linearity and noise sources in flow cytometry.

A model of pulse formation in flow cytometers is presented that demonstrates the proportionality between the area (or the peak height) of the fluorescence signal produced by the photomultiplier and the number of fluorochrome molecules present in the cell that cause the signal. The model clarifies the possible instrumental origins of inaccuracy in this linearity that results in a broadening of the histograms obtained. A comprehensive formula for the coefficient of variation of the unimodular histograms of an homogeneous population is presented that clearly discriminates among the different contributions of staining, possible inhomogeneity of the examined population, photon statistics, and instrumental instabilities. Finally, some experimental data are presented that show the agreement with the proposed formula.