High-speed fluorescence lifetime imaging

Abstract

In Fluorescence Lifetime Imaging (FLIM) the ns fluorescence decay is used for imaging. It yields information about the local environment of the fluorescent molecule and is very well suited for quantitative imaging. FLIM was introduced more than one decade ago now. Most of the implementations of FLIM require comparatively long acquisition times on the order of ten seconds or more. This hampers the use of FLIM for the study of dynamic processes. In FLIM at least one order of magnitude more signal is required than in conventional intensity imaging. Therefore, fast FLIM acquisition rates require efficient detection schemes and detectors. We evaluated the count rate limitation in time gated and TCSPC FLIM of a number of single photon counting detectors. In particular we looked at the performance of a conventional fast head-on PMT (R1894), a GaAs photocathode PMT (H7422P-40) and a single photon counting avalanche photo diode (SPCM-AQR14). The high quantum efficiency GaAs photocathode PMT and avalanche photo diode detectors show lifetime shifts in both time gated detection and in TCSPC starting at a detection count rate of 1 - 2 MHz. The conventional PMT shows lifetime shifts starting at a detection count rate of about 2.5 MHz in TCSPC and at about 6 MHz for the time gated detection system. The detection efficiency of the TCSPC based system goes down rapidly above about 1 MHz due to the dead time of the detection electronics. The time gated detection system shows little or no reduction of the detection efficiency up to detection count rates of 10 MHz with the conventional fast PMT. The time gated-detection system was coupled to a multi-photon excitation microscope. Calcium transients were recorded in cardiac rat myocytes at a 1 Hz frame rate. The system operated at the full repetition rate of the Ti:Sa laser. Here, the frame rate was limited by the maximum count rate of the PMT.