Two-photon fluorescence lifetime imaging microscopy utilizing the space-time duality

PROJECT SUMMARY Fluorescence lifetime imaging microscopy (FLIM) is a type of fluorescence imaging technologies that is gaining popularity in biomedicine because it delivers the most direct insight into the molecular conformation and the biological environment of a fluorophore. FLIM has been applied to provide insights into the cellular metabolism, protein-protein interactions, and biological environment monitoring of temperature, viscosity, pH, and ion concentration. Despite the wealth of information provided by the FLIM, its widespread application is currently limited by the low imaging speed. The FLIM imaging speed is a complex function of many factors, with shot noise by the photon counting statistics being the fundamental limit. This limitation is especially dominant for fluorophores with lifetime shorter than the FLIM instrument response function (IRF) when deconvolution is necessary to accurately determine the fluorescence lifetime. Thus, to fundamentally enhance the FLIM imaging speed, either an increase of the maximum photon counting rate or a reduction of the FLIM IRF is necessary. Time-domain FLIM with high photon efficiency can be implemented with either time-correlated single-photon counting (TCSPC) or photon counting streak camera (PCSC). The maximum photon counting rate of state-of- the-art time-domain FLIM is 1-10 mega counts per second (Mcps), limited by the pile up effect in TCSPC-FLIM and the readout nonlinearity and crosstalk in PCSC-FLIM. TCSPC-FLIM generally has a 100-ps IRF, unless superconducting nanowire single-photon detectors that require cryogenic cooling are implemented to reach the picosecond regime. On the other hand, PCSC-FLIM can achieve the picosecond IRF at room temperature, but complex streaking and detection optoelectronics are required. Using PCSC-FLIM, a recent study on Alzheimer mouse brain tissue has found a new 30-ps lifetime component, critical for separating Alzheimer disease from normal brain tissue, of nicotinamide adenine dinucleotide hydrate (NADH). Without the 10-ps IRF of PCSC-FLIM, such fast fluorescence decay could not have been observed within a reasonable amount of time. Similarly, a short IRF will benefit the study of short-lived non-lipofuscin autofluorophores (30-70 ps) that will lead to a better understanding of the fundus autofluorescence diagnosis and may provide relevant retina information for the early detection of age-related macular degeneration and neurodegenerative diseases. This proposal will develop a potentially transformative FLIM system, photon-streaking FLIM (PS-FLIM), that addresses the imaging speed challenge by simultaneously reducing the IRF and increasing the maximum photon counting rate. A new concept of photon streaking, based on the principle of space-time duality, will be implemented to achieve 5-ps IRF and 840 Mcps in a compact and lightweight platform. Two-photon excitation will be utilized to increase the imaging depth and reduce the phototoxicity. Finally, machine learning framework will be incorporated to accelerate the FLIM data analysis.