Fluorescence lifetime imaging (FLI) aims at quantifying the exponential decay rate of fluorophores to yield lifetime maps over the imaged sample. When combined with Forster resonance energy transfer (FRET), the technique can be used to indirectly sense interactions at the nanoscale such as protein-protein interactions, protein-DNA interactions, and protein conformational changes. In the case of FLI-FRET, the fluorescence intensity decays are fitted to a bi-exponential model in order to estimate the lifetime and fractional amplitude coefficients of each component of the population of the donor fluorophore (quenched and non-quenched). Numerous time data points, also called temporal or time gates, are typically employed for accurately estimating the model parameters, leading to lengthy acquisition time and significant computational demands. This work investigates the effect of the number and location of time gates on model parameter estimation accuracy. A detailed model of a FLI-FRET imaging system is used for the investigation and simulation outcomes are validated with in vitro and in vivo experimental data. In all cases investigated, it is found that 10 equally spaced time-gates allows robust estimation of model-based parameters with similar accuracy to full temporal data sets (90 gates).
Journal of Biomedical Optics 19, No. 8, 086023 (2014)