Chiral drugs play an important role in modern medicine, but obtaining pure enantiomers from racemic mixtures can pose challenges. When a drug is chiral, only one enantiomer (eutomer) typically exhibits the desired pharmacological activity, while the other (distomer) may be biologically inactive or even toxic. Racemic drug formulations introduce additional health risks, as the body must still process the inactive or detrimental enantiomer. Some distomers have also been linked to teratogenic effects and unwanted side effects. Therefore, developing efficient and scalable methods for separating chiral drugs into their pure enantiomers is critically important for improving patient safety and outcomes. Metal-organic frameworks (MOFs) show promise as novel materials for chiral separation due to their highly tunable structures and interactions. This review summarizes recent advancements in using MOFs for chromatographic and spectroscopic resolution of drug enantiomers. Both the opportunities and limitations of MOF-based separation techniques are discussed. A thorough understanding of these methods could aid the continued development of pure enantiomer formulations and help reduce health risks posed by racemic drug mixtures.
Despite growing clinical adoption of artificial intelligence in radiology, fewer than one in five commercially available models are cleared for paediatric use. Existing systems are trained predominantly on adult data and generalize poorly to neonates, yet efficient strategies for repurposing them without data sharing or large-scale local data curation are lacking. Here we present an international retrospective study of neonatal chest radiographs from 30 neonatal intensive care units across 20 countries and territories, evaluating continual learning as a privacy-preserving approach to adapt a deep learning model for adult endotracheal tube assessment to neonatal care. Each hospital contributed 50 radiographs for sequential model retraining without centralized data pooling. Continual learning achieved lower anatomical landmark localization error (0.67 cm versus 1.29 cm for hospital-specific fine-tuning and 4.30 cm for adult-only models) at 29 of 30 sites, with improved detection of deep intubations at 27 of 30 sites. These findings demonstrate that privacy-preserving collaborative learning can address critical gaps in paediatric AI deployment.