Cécile Darviot, Michel Meunier
Hyperspectral imaging combined with plasmonic nanoparticles offers powerful opportunities for nanoscale biological diagnostics, yet conventional push-broom systems are limited by slow acquisition and reduced spatial resolution. We report a wide-field hyperspectral imaging platform integrating a custom linear variable tunable filter (LVTF) with reflected light microscopy (RLM) for multiplexed plasmonic nanoparticle (NP) imaging. While LVTF-based systems have been used in industrial contexts, their adaptation to RLM microscopy enables high-resolution, high-contrast spectral imaging of NPs on biological samples. The system acquires data across 380-720 nm with 7-20 nm spectral resolution while preserving spatial detail. We demonstrate imaging of NPs bound to cancer cell membranes and classification of four distinct NP types. A trained support vector machine achieves an overall F-score of ~97%, significantly outperforming RGB-based approaches. Compared to push-broom systems, this platform provides faster acquisition, improved spatial resolution, and a compact, cost-effective solution for multiplex NP detection in bioimaging and immunoplasmonic applications.