High-speed lifetime imaging enables autofluorescence-free immunofluorescence for quantitative pathology

Endogenous autofluorescence (AF) often interferes with antibody-conjugated fluorophore emissions, reducing the sensitivity and accuracy of immunofluorescence (IF) microscopy. Existing methods such as chemical quenching, targeted photobleaching, and digital subtraction show limited efficacy across diverse biochemical and morphological clinical specimens. In this study, we utilized a GPU-accelerated FLIM capable of visualizing lifetime maps in real-time. Due to the distinct phasor positions of antibody-conjugated fluorophores compared to common autofluorescence biomolecules, pixel-wise phasor unmixing enabled effective real-time separation of specific IF signals from background autofluorescence. Through quantitative Monte Carlo simulations, we demonstrated that approximately 98% autofluorescence suppression is achievable when photon counts per pixel reach around 900. These results were validated experimentally using human formalin-fixed paraffin-embedded (FFPE) tissues. In short, This study shows high speed FLIM as a promising technique to streamline quantitative pathology and high-content screening workflows.