Recent research shows that the mechanical response of human brain tissue depends strongly on the testing modality of the experiments, with notable discrepancies between in-vivo and ex-vivo measurements. These differences highlight the need for predictive computational models that can reconcile these diverse experimental observations and capture key physiological features that cannot be replicated ex-vivo. In this work, we extend a viscoelastic brain tissue model by explicitly incorporating the cerebral vasculature and associated arterial pressure. The objective is to account for solid-fluid interactions between the brain tissue and the vascular network, which are absent in ex-vivo experiments and may significantly influence in-vivo mechanical responses. To achieve this, we employ a multiscale immersed method [1] in which blood vessels are embedded within the continuum brain tissue domain. Vascular pressure is introduced as a forcing term, and coupling between the vascular and tissue scales is enforced using a reduced Lagrange multiplier formulation. The proposed framework is applied to perform magnetic resonance elastography (MRE) in-silico. Previous in-vivo MRE studies have shown that cerebral blood flow substantially affects measured viscoelastic properties [2], suggesting that vascular loading may even induce an effective pre-stress in the tissue. By comparing material parameters recovered from inverting simulated displacements with and without vascular integration, we assess the influence of vascular pressure on the apparent mechanical behaviour of brain tissue. Our results demonstrate that inclusion of the vascular network leads to measurable stiffening in the estimated viscoelastic parameters, providing a mechanistic explanation for discrepancies between in-vivo and ex-vivo measurements [3]. This work underscores the importance of vascular contributions in brain biomechanics and offers a modelling framework for more physiologically interpretation of in-vivo experimental data.