Neuro-multimodal monitoring in neonatal brain injury: Current applications, challenges, and future directions

Advances in neuromultimodal monitoring (MMM) have significantly enhanced early detection and prognostic evaluation of brain injury in neonatal intensive care. This review summarizes the application and integration of multimodal monitoring technologies for brain injury assessment in newborns. Compared to single-modality approaches, MMM-based combined monitoring strategies demonstrate superior accuracy in both injury identification and prediction of neurodevelopmental outcomes. Meanwhile, artificial intelligence (AI) methods are increasingly being utilized for multi-source data integration and modeling, showing considerable potential in identifying subtle biomarkers and optimizing risk stratification.However, several challenges—including insufficient standardization of monitoring parameters, lack of device calibration protocols, limited multicenter data sharing, and the "black-box" nature and poor clinical interpretability of AI algorithms—hinder the clinical translation and broad application of MMM technologies. Looking forward, further development of MMM requires the establishment of unified standardized frameworks and open-access databases, alongside enhanced validation of AI-assisted algorithms for transparency and reliability, and the development of more efficient data integration models. These advances are essential to solidify the core role of MMM in neonatal neuromonitoring systems. • Multimodal monitoring (MMM) integrates electroencephalography (EEG), near-infrared spectroscopy (NIRS), transcranial Doppler (TCD), and neuroimaging to provide a comprehensive, real-time assessment of neonatal brain injury, overcoming the limitations of single-modality approaches. • Artificial intelligence (AI) enhances MMM through advanced data integration and predictive modeling of neurodevelopmental outcomes, though clinical translation remains challenged by issues of interpretability and standardization. • Key techniques such as amplitude-integrated EEG (aEEG) for seizure detection, NIRS for cerebral oxygenation, and TCD for cerebral perfusion demonstrate high sensitivity in early injury identification and prognostic stratification. • Gestational age-specific reference values and individualized hemodynamic management are essential for the accurate interpretation of neuromonitoring parameters in both preterm and term infants. • Future development requires standardized protocols, open-access databases, and clinically transparent AI models to facilitate the integration of MMM into routine neonatal care and improve long-term neurodevelopmental outcomes.