Calibrating the BOLD signal revisited – Calculation of oxygen metabolism for gradient- and spin-echo sequence up to 16.4T

Introduction The baseline fMRI signal and the blood oxygenation level-dependent (BOLD) signal amplitude are not a quantitative reflection of neuronal activity as physiological and physical parameters (e.g. baseline CBF, echo time, coil sensitivity ...) contribute to the the signal. One goal of quantitative fMRI is to determine oxygen metabolism (CMRO2) from fMRI data. To this end, a calibrated BOLD approach has been proposed calculating oxygen metabolism from combined CBF and BOLD data [1,2]. The calibrated BOLD approach for GRE sequence at 1.5T has been derived under the assumption that only extra-vascular signal from mostly veins contributes to the BOLD signal. Recently, we have proposed an alternative model, named the ‘integrative model’ which takes into account intraand extravascular contributions of microand macro-vasculature for both GRE and SE up to 16.4T [3]. Using the integrative model, a general calibration model of the BOLD signal as a function of oxygen extraction fraction (OEF) and total CBV is derived and applied, for the first time, on 7T GRE and SE fMRI data obtained on the macaque monkey. Methods For simplicity, the total BOLD signal as calculated from the integrative model [3] for GRE respectively for SE is fitted with a first order equation, which is our main equation to determine CMRO2 (see table for macro-vasculature):