The abstract is read at the publisher; the summary is JClub's.
Vision Transformer–based hyperspectral stimulated Raman scattering microscopy accurately and non-destructively classifies multiple lymphocyte phenotypes, including T and B lineage and T-cell states, using label-trained attention features.
- Why it matters: Immune profiling needs both subtype and functional state without destructive endpoint assays or complex labeling that can perturb cell behavior.
- What they did: The method uses attention-based spatial–spectral feature learning on hyperspectral SRS images, with fluorescent markers only for training to label T and B lineages, T subsets, differentiation (effector, memory, naive), and functional states (naive, activated, exhausted).
- The result: The approach preserves cells for downstream or longitudinal work and enables label-free immune profiling and live-cell analysis of type and state transitions.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.