Performance of two clinical fluorescence imaging systems with different targeted and non-targeted near-infrared fluorophores: a cadaveric explorative study

Introduction: Near-infrared (NIR) fluorescence-guided surgery is increasingly utilized in humans and pets. As clinical imaging systems are optimized for Indocyanine green (ICG) detection, the usage of targeted dyes necessitates the validation of these systems for each dye. We investigated the impact of skin pigmentation and tissue overlay on the sensitivity of two NIR cameras (IC-Flow^TM, Visionsense^TM VS3 Iridum) for the detection of non-targeted (ICG, IRDye800) and targeted (Angiostamp^TM, FAP-Cyan) NIR fluorophores in an ex vivo big animal model.

Methods: We quantitatively measured the limit of detection (LOD) and signal-to-background ratio (SBR) and implemented a semi-quantitative visual score to account for subjective interpretation of images by the surgeon.

Results: Visionsense^TM VS3 Iridum outperformed IC-Flow^TM in terms of LOD and SBR for the detection of all dyes except FAP-Cyan. Median SBR was negatively affected by skin pigmentation and tissue overlay with both camera systems. Level of agreement between quantitative and semi-quantitative visual score and interobserver agreement were better with Visionsense^TM VS3 Iridum.

Conclusion: The overlay of different tissue types and skin pigmentation may negatively affect the ability of the two tested camera systems to identify nanomolar concentrations of targeted-fluorescent dyes and should be considered when planning surgical applications.