Abstract
Significance: The cerebral metabolic rate of oxygen ( CMRO2 ) is an important indicator of brain function and pathology. Knowledge about its magnitude is also required for proper interpretation of the blood oxygenation level-dependent (BOLD) signal measured with functional MRI. Despite the need for estimating CMRO2 , no gold standard exists. Traditionally, the estimation of CMRO2 has been pursued with somewhat indirect approaches combining several different types of measurements with mathematical modeling of the underlying physiological processes. The recent ability to measure the level of oxygen ( pO2 ) in cortex with two-photon resolution in in vivo conditions has provided a more direct way for estimating CMRO2 , but has so far only been used to estimate the average CMRO2 close to cortical penetrating arterioles in rats. Aim: The aim of this study was to propose a method to provide spatial maps of CMRO2 based on two-photon pO2 measurements. Approach: The method has two key steps. First, the pO2 maps are spatially smoothed to reduce the effects of noise in the measurements. Next, the Laplace operator (a double spatial derivative) in two spatial dimensions is applied on the smoothed pO2 maps to obtain spatially resolved CMRO2 estimates. Result: The smoothing introduces a bias, and a balance must be found where the effects of the noise are sufficiently reduced without introducing too much bias. In this model-based study, we explored this balance using synthetic model-based data, that is, data where the spatial maps of CMRO2 were preset and thus known. The corresponding pO2 maps were found by solving the Poisson equation, which relates CMRO2 and pO2 . MATLAB code for using the method is provided. Conclusion: Through this model-based study, we propose a method for estimating CMRO2 with high spatial resolution based on measurements of pO2 in cerebral cortex.