Oxygen availability in brain tissue is closely linked to local hemodynamics and even slight disturbances in the cerebral microcirculation may damage cells due to the brain’s high energy demands. In addition to local cerebral blood flow, knowledge of the oxygen extraction fraction (OEF) is critical when assessing brain tissue oxygenation. A biophysical model that relates the brain’s microvascular hemodynamics to OEF has previously been proposed. Here, we aimed to calibrate and compare this model with OEF measurements determined by [ 15 O]-based positron emission tomography imaging (PET). Local brain hemodynamics were assessed in 68 healthy elderly individuals using dynamic susceptibility contrast magnetic resonance imaging (DSC-MRI). Average DSC-MRI-based mean transit time and capillary transit time heterogeneity were compared to PET OEF to calibrate the model parameters. The calibrated biophysical model produced OEF estimates in the range of PET OEF with a moderate correlation (r = 0.31, p = 0.009), albeit with a tendency to overestimate smaller PET OEF values and underestimate larger PET OEF values. We discuss the assumptions made when modeling oxygen transport in measurements of local hemodynamics and in [ 15 O]-based tracer uptake, respectively, and propose that the biophysical model provides a valuable tool to link hemodynamic changes to oxygen uptake in the human brain.

Conventional MRI is crucial for diagnosing multiple sclerosis (MS) but lacks precision, leading to the clinico-radiological paradox and misdiagnosis risk, especially when confronted with unspecific lesions not related to MS. Advancements in perfusion-weighted imaging (PWI) with an algorithm designed for diseases with anticipated contrast agent extravasation offer insight into microvascular impairment and flow heterogeneity. Our study aimed to assess these factors in MS patients and their association with clinically relevant white matter injury and disease course. We evaluated 60 adults with white matter lesions (WML), including 50 diagnosed with MS or MS syndromes and 10 non-diseased symptomatic controls (SC) with unspecific WML. MRI included conventional three-dimensional (3D) T2-weighted fluid-attenuated inversion recovery (T2-FLAIR), 3D magnetization-prepared two rapid acquisition gradient-echo (MP2RAGE), post-contrast 3D T1-weighted (T1) images, and Dynamic Susceptibility Contrast (DSC) PWI at 3T. WML masks of “unspecific T2-FLAIR lesions”, “MS T2-FLAIR lesions”, and “MS T1-lesions” were manually outlined and validated by a neuroradiologist. DSC-derived parameters were analyzed in WML masks and healthy-appearing tissue. MS T2-FLAIR lesions showed increased flow heterogeneity and vasodilation compared to unspecific T2-FLAIR lesions in SC, as well as compared to unspecific T2-FLAIR lesions within the MS group. MS T1-lesions exhibited more homogenized flow. Our findings suggest that DSC-PWI, combined with lesion delineation, can provide clinically relevant differentiation of MS lesions from unspecific WML, highlighting potential microvascular pathology previously overlooked in MS.