Current modeling approaches for drug delivery to the central nervous system

Central nervous system (CNS) disorders pose a major global health challenge, yet therapeutic development is impeded by the difficulty of delivering effective drug concentrations to the brain. Based on a literature search of PubMed, Scopus, and Google Scholar (1990-2025), this review delineates the current landscape of computational modeling techniques addressing CNS drug delivery, emphasizing anatomical barriers and physiological transport mechanisms relevant to major neurological diseases. We categorize approaches spanning the molecular dynamics interactions of drug-blood-brain barrier (BBB) to macroscopic continuum and physiologically based pharmacokinetic (PBPK) models that elucidate systemic distribution and brain exposure. These models are assessed across established delivery routes, such as intranasal and intrathecal administration, and emerging methods, including focused ultrasound-mediated BBB opening and targeted nanoparticle delivery. We highlight the growing importance of integrating complex physiological phenomena, such as glymphatic flow and cerebrospinal fluid (CSF) dynamics, into predictive models. Finally, we explore opportunities involving multiscale digital twins of the CNS that integrate molecular interactions, vascular hemodynamics, perivascular flow, and parenchymal transport within patient-specific geometries. We also examine the role of machine learning and surrogate modeling in accelerating prediction of drug transport parameters and optimizing delivery strategies, aiming to guide the design of robust computational platforms.