Single-entity analysis of therapeutic nanoparticles: from orthogonal metrology to AI

Therapeutic nanoparticles, including lipid nanoparticles (LNPs), polymeric nanocarriers, and inorganic nanomaterials, are intrinsically heterogeneous in size, mass, composition, and drug loading. However, their characterization still relies largely on ensemble-averaged methods, which obscure particle-to-particle variability and overlook rare but functionally important subpopulations. This limitation has become a major obstacle to understanding structure - function relationships, ensuring batch reproducibility, and supporting translational development. This review highlights emerging strategies for single-entity analysis of therapeutic nanoparticles, emphasizing the shift from bulk-average measurements to particle-resolved characterization. We first outline the multidimensional information accessible at the single-particle level and explain why no single technique can fully capture nanoparticle heterogeneity. We then summarize recent advances in orthogonal single-particle measurements, including mass-, element-, composition-, and separation-based approaches. In addition, we discuss the growing role of Artificial Intelligence (AI) - enabled multimodal analysis in automated signal processing, subgroup identification, and cross-platform data integration. Together, these advances provide a more informative and reliable framework for therapeutic nanoparticle characterization. This narrative review was based on literature searches of PubMed, Web of Science, and Google Scholar for articles published from January 2014 to March 2026.