Perinatal brain injury (PBI) is a leading cause of childhood morbidity and mortality, often resulting in long-term neurological deficits. Despite its relatively high prevalence, PBI lacks reliable biomarkers for early detection and effective therapeutic approaches. The etiology of PBI is multifactorial and includes not only preterm birth but also hypoxia-ischemia, infection, and inflammation. While animal models and two-dimensional cell cultures have contributed to our understanding of PBI pathophysiology as well as to the development of novel treatment strategies, they fail to fully capture the complexity of injury in the developing human brain. Human brain organoids have recently emerged as transformative platforms recapitulating key features of fetal-neonatal brain development, including relevant cell types, gene expression patterns, cytoarchitecture, and functional electrophysiological properties. This model therefore provides opportunities to investigate the mechanisms underlying perinatal insults and serves as a promising tool for novel treatments such as stem cell-based therapy, drug discovery and screening, helping to bridge the translational gap between preclinical studies in animal models and clinical applications. Recent innovations, including the development of vascularization strategies, the incorporation of glial cells, brain organoid-on-chip technologies, and assembloids, have further increased the physiological relevance of these methods. This review highlights recent advances in organoid-based models of PBI and highlights their potential and challenges as next-generation tools for mechanistic studies and therapeutic innovations.
Zahra Dehghani, D. Surbek, M. Joerger-Messerli et al.· Stem Cell Research & The...· 0 citations
Training fast generators on new domains with limited data remains challenging for two reasons. First, adapting a pretrained diffusion or flow model to a new domain leaves its costly multi-step sampling unaddressed, and existing acceleration methods are tied to the source parameterization--$\epsilon$, $x$, $v$, or $u$--leaving heterogeneous pretrained models with no common acceleration target. Second, while adversarial refinement is proven effective for few-step quality, it is formulated only for instantaneous-velocity flows, not for the finite-interval average velocities that MeanFlow (MF) models predict. We address both problems. We propose MeanFlow-Transfer, which maps heterogeneous source outputs into a shared velocity representation, uses it to initialize an MF generator from the source weights, and optimizes an MF objective on the target domain. This unifies adaptation and acceleration in a single training loop across a broad range of pretrained models. We then introduce Continuous Adversarial MeanFlow, a post-training stage that extends continuous adversarial flow models from instantaneous velocities to MF's finite-interval average velocities. CAMF contrasts changes in a learned potential between real and predicted interval endpoints, recovering fine detail that MF regression averages away, and reduces to the instantaneous criterion in the vanishing-interval limit. Adapting four ImageNet-based source models--DiT ($\epsilon$), SiT ($v$), JiT ($x$), iMF ($u$)--to five target domains, MF-T with CAMF matches or exceeds the fine-tuned teacher in FID and FDD at up to $125\times$ fewer Neural Function Evaluations (NFEs), while CAMF improves MF-T's few-step FID by $29\%$ on average.
Yara Bahram, Zahra Dehghani, M. Desbos et al.· 0 citations