Distribution and reliability of multi-parametric quantitative fMRI across the human cortex

Multi-parametric quantitative MRI is a technique designed to quantify parameters related to oxygen metabolism in the brain, both during task performance and at baseline. In this thesis, mqBOLD resting-state data from four datasets, acquired using two different scan- ners, were analyzed. Parameter maps of the cerebral metabolic rate of oxygen (CMRO2), oxygen extraction fraction (OEF), cerebral blood flow (CBF), and cerebral blood volume (CBV) from healthy subjects were examined. Statistical analysis revealed no significant differences in the distributions of these parameters within data acquired using the same scanner, although inter-scanner differences were observed. Despite these differences, all four datasets demonstrated consistent metabolic patterns, and the frontoparietal network was identified as the highest oxygen-consuming network, with an average CMRO2 value of 157.4 µmol/100g/min. Furthermore, intraclass correlation analysis on subjects scanned multiple times revealed excellent CMRO2 reliability within the same session (ICC > 0.9), but only moderate reliability between different days (0.4 < ICC < 0.7), while CBF and CBV showed higher values in both cases (ICC > 0.6). Partial least squares analysis in- dicated higher CMRO2 (t -statistic = −2.99, p = 0.004) and OEF (t -statistic = −3.41, p = 0.002) values in women, compared to men. Additionally, age was found to be pos- itively correlated with OEF (r = 0.36, p = 0.001) and negatively correlated with CBF (r = −0.36, p = 0.001). These findings contribute to the understanding of healthy brain metabolism at rest, providing a reliable quantitative baseline for comparison, advanc- ing the knowledge of the neural and metabolic underpinnings of the BOLD signal, and establishing criteria to distinguish between health and disease.