Patient-specific arterial input function for accurate perfusion assessment in intraoperative fluorescence imaging.

Significance: The arterial input function (AIF) plays a crucial role in correcting the time-dependent concentration of the contrast agent within the arterial system, accounting for variations in agent injection parameters (speed, timing, etc.) across patients. Understanding the significance of the AIF can enhance the accuracy of tissue vascular perfusion assessment through indocyanine green-based dynamic contrast-enhanced fluorescence imaging (DCE-FI).

Aim: We evaluate the impact of the AIF on perfusion assessment through DCE-FI.

Approach: A total of 144 AIFs were acquired from 110 patients using a pulse dye densitometer. Simulation and patient intraoperative imaging were conducted to validate the significance of AIF for perfusion assessment based on kinetic parameters extracted from fluorescence images before and after AIF correction. The kinetic model accuracy was evaluated by assessing the variability of kinetic parameters using individual AIF versus population-based AIF.

Results: Individual AIF can reduce the variability in kinetic parameters, and population-based AIF can potentially replace individual AIF for estimating wash-out rate ( $k_{\mathrm{ep}}$ ), maximum intensity ( $I_{\max }$ ), ingress slope with lower differences compared with those in estimating blood flow, volume transfer constant ( $K^{\mathrm{trans}}$ ), and time to peak.

Conclusions: Individual AIF can provide the most accurate perfusion assessment compared with assessment without AIF or based on population-based AIF correction.