Use of retinal oximetry in estimating cerebral tissue oxygenation
Abstract
Purpose
To investigate the correlation between cerebral tissue oxygenation (SctO2) and retinal vessel oxygen saturation (SO2) using non-invasive spectroscopy in healthy individuals.
Methods
Retinal and cerebral oxygen saturations were measured in dark-adapted, healthy volunteers breathing ambient air in a seated position using dual wavelength retinal oximetry and transcranial near-infrared spectroscopy (NIRS) respectively. Correlations between SO2 and SctO2 were analyzed using Pearson correlation coefficients. Multivariate analysis was performed to determine the relative contribution of the arterial and venous vessels to SctO2. Using this model, SctO2 was estimated based on retinal arterial and venous oxygen saturation. Pearson correlation coefficients, paired sample t-test and Bland-Altman analysis were used to assess the agreement between the measured and the predicted SctO2.
Results
Twenty-one young healthy individuals aged 26.4 ± 2.2 years were analyzed. SctO2 showed a significant positive correlation with both arterial and venous SO2 (r = 0.442, P = 0.045 and r = 0.434 P = 0.049 respectively). In multivariate analysis, the relative contribution of arterial and venous SO2 to SctO2 was significantly correlated with diastolic blood pressure, retinal venous oxygen saturation and retinal venous diameter (R2=0.60, P < 0.001). The measured SctO2 (72.2 ± 3.5%, range 67.3–80.1) correlated well with estimated SctO2 (72.2 ± 2.6%, range 67.6–77.5) (r = 0.774, P < 0.001). Bland–Altman plots showed 95% agreement within ±6.8%.
Conclusions
In this pilot study, retinal oximetry showed promising as an estimate of cerebral tissue oxygenation as measured by NIRS.
