Normalization is essential for accurate intraoperative perfusion assessment using near-infrared fluorescence imaging

<i>Significance</i>. Near-infrared fluorescence (NIRF) imaging is increasingly used for perfusion assessment in clinical care and research settings. While subjective interpretation demonstrates clinical benefits, quantitative analysis is crucial for broader adoption and classification of perfusion patterns. However, strict standardization often conflicts with fast-paced clinical workflow, obstructing broad implementation.<i>Aim and approach</i>. This study hypothesized that normalization of fluorescence measurements could effectively correct for measurement variability, reducing the need for strict standardization in quantitative NIRF perfusion imaging. To evaluate this, a model capable of consistently simulating fluorescence perfusion patterns was employed. Experiments were conducted in an operating room using four NIRF camera systems under varying measurement conditions.<i>Results</i>. Normalization to maximum signal intensity provided consistent perfusion parameter values across differences in camera type, angle, rotation and settings (Δ≤1%). In cases of severe malperfusion where peak-intensity was not reached within the measurement period, normalization did not adequately correct for measurement variability (Δ increased).<i>Conclusion</i>. While standardization remains valuable beyond parameter accuracy, appropriate normalization substantially reduces dependence on strict measurement protocols. These findings support broader clinical adoption of quantitative NIRF imaging, extending its utility beyond specialized tertiary centers and facilitating widespread integration into routine surgical care.