Label-free imaging reads immune cell activity from routine blood samples with 94% accuracy

A new label-free approach based on autofluorescence lifetime imaging was applied to peripheral blood mononuclear cells to assess their metabolic state at the single-cell level. Using two-photon microscopy, researchers measured the natural light emitted by intracellular molecules and fed the lifetime data into machine-learning models. Tests on samples from three donors showed the method could separate activated immune cells from resting ones with nearly 94% accuracy just two hours after stimulation.

It also identified monocytes with up to 96% accuracy in resting samples and 88% after activation, while NK cells were recognized with about 74% accuracy. The study highlighted metabolic signatures as an additional layer of information beyond conventional surface markers, revealing heterogeneity within cell populations that bulk assays would miss. Although not intended to replace traditional fluorescent labeling, the technique offers a non-destructive way to gauge cell function, potentially aiding research on cancer, immune disorders, and cell-based therapies.