Continuous assessment of tissue perfusion using quantitative indocyanine green fluorescence imaging during controlled...

Indocyanine green fluorescence imaging (ICG-FI) is a novel tool for continuous assessment of tissue perfusion. However, consensus on the optimal methodology is lacking.  It was hypothesized that quantitative ICG-FI parameters based on continuous micro-dosing correlate with corresponding transit time flow measurements in the functional renal end artery.  This study aimed to examine the feasibility of continuous, quantitative ICG-FI for detecting perfusion changes (hypoperfusion, reperfusion) in porcine kidneys.

Renal perfusion was assessed using continuous, quantitative ICG-FI under controlled renal artery flow adjustments (no, partial, full, partial, and no occlusion) in ten healthy female pigs. Four 0.008 mg/kg ICG micro-doses were administered with 60-s intervals for each flow adjustment. As a reference, simultaneous renal artery transit time flow probe measurements were recorded. ICG-FI parameters were extracted by PerfusionWorks^®, providing surrogate markers for perfusion, and correlated to the reference renal artery transit time flow measurements using linear regression modeling.

During hypoperfusion, mean flow decreased from 277 to 139 to 0 ml/min in the no, partial, and full occlusion steps, respectively. F_ingress (R² = 79%), F_max (R² = 79%), and slope (R² = 78%) correlated with flow. During reperfusion, mean flows increased from 0 to 169 to 240 ml/min in the full, partial, and no occlusion steps, respectively, and F_ingress (R² = 71%), F_max (R² = 69%), and slope (R² = 79%) also correlated with flow. Overall, slope exhibited the strongest correlation (R² = 54%). T_ingress, T_½max, time ratio, and T_max exhibited poor correlations with flow.

Continuous assessment of renal perfusion was technically feasible. In this kidney model, the combined ICG-FI parameter, slope, and the intensity-dependent parameters F_ingress and F_max exhibited convincing correlation with flow, appearing promising for the continuous assessment of tissue perfusion. Time-related parameters did not correlate with flow in this specific model, likely due to the kidney's unique vascular characteristics.