A research team from Tsinghua University has introduced a breakthrough method that fundamentally reduces photon requirements for fluorescence lifetime imaging microscopy, enabling high-fidelity, low-light biological imaging. Led by Academician Dai Qionghai and Associate Professor Wu Jiamin, with lead author Zhou Yiliang, the work was recently published in Nature Biotechnology. The team developed EFLIM, an event-based first-photon FLIM technique that overcomes a longstanding bottleneck in quantitative bio-imaging: the reliance on thousands of photons per pixel to generate reliable lifetime measurements. Traditional FLIM captures how long a fluorophore remains in an excited state before emitting a photon, providing critical insights into cellular microenvironments, ion concentrations, and molecular interactions independent of signal intensity. However, precise lifetime estimation traditionally requires accumulating tens of thousands of photon arrival times into a histogram. This process slows acquisition, increases phototoxicity, and degrades rapidly in deep or scattering tissues. EFLIM circumvents this by abandoning histogram accumulation entirely. Instead, it treats each laser excitation pulse as an independent event, extracting lifetime information solely from the arrival time of the very first photon detected per pixel. The system viability in ultra-low-photon regimes hinges on an AI-driven spatiotemporal self-supervised denoising framework. Rather than relying on paired high-photon ground-truth data, the model leverages statistical consistency across adjacent pixels and consecutive frames to reconstruct accurate lifetime estimates from sparse photon events. The researchers calibrated performance boundaries through extensive simulations and empirical validation, confirming that EFLIM maintains quantitative reliability even when photon budgets drop below one per pixel per frame. Experimental demonstrations across multiple biological models underscore the method practical impact. In live HeLa cells, EFLIM captured fast calcium dynamics at thirty frames per second under extreme photon limitation. Awake mouse brain imaging revealed stable neuronal lifetime readings at depths of 250 micrometers, effectively mitigating motion-induced artifacts. In immunological assays, the technique differentiated B and T cells within a single spectral channel and tracked intercellular vesicle communication. Additionally, EFLIM enabled rapid, label-free mapping of human glioma tissues, distinguishing tumor margins, vasculature, and necrotic zones with clinical relevance. By decoupling fluorescence lifetime quantification from high photon counts, the approach expands the viability of weak or low-expression probes, facilitates deep-tissue neuroimaging, and accelerates label-free histopathology. Future iterations aim to integrate higher quantum-efficiency single-photon detectors and combine lifetime data with spectral and polarization channels for multiplexed molecular profiling. The work establishes a new paradigm for quantitative biology, transforming FLIM from a low-throughput, high-photon technique into a rapid, minimally invasive tool for dynamic in vivo and clinical applications.