Translating NIR-II imaging into clinical practice: advances and strategies

Near-infrared II (NIR-II) bioimaging has advanced rapidly through the development of diverse luminescent probes, improved detection hardware, and increasingly sophisticated imaging modalities. However, the translation of NIR-II imaging from preclinical demonstrations to practical biomedical use remains limited by a central challenge: probe performance, imaging-system configuration, computational processing, and application requirements are often optimized separately rather than as an integrated system. In this review, we discuss NIR-II bioimaging from a task-driven and system-level perspective. We first summarize the optical basis of NIR-II imaging and the evolution of major probe families, including single-walled carbon nanotubes (SWCNTs), quantum dots (QDs), organic fluorophores, rare earth nanoparticles (RENPs), and metal nanoclusters. We then analyze how excitation sources, optical architectures, detectors, lifetime strategies, and computational reconstruction methods shape the effective performance of NIR-II imaging in vivo. Representative applications in tumor surgery, vascular and lymphatic imaging, brain imaging, and other disease models are further discussed according to their translational evidence levels. Building on these analyses, we propose a closed-loop framework in which clinical tasks define the required imaging window, probe properties, system architecture, quantitative readouts, and validation criteria. This framework emphasizes that future NIR-II bioimaging should move beyond isolated improvements in wavelength or brightness toward standardized, application-matched, and clinically realistic co-design of probes, instruments, algorithms, and biomedical workflows.