Precision neuroprotection in cardiac surgery: a structured narrative review of multimodal, data-driven perioperative care

Neurological injury remains a major contributor to morbidity, mortality, and long-term cognitive decline in patients undergoing cardiac surgery, despite significant advancements in surgical techniques and cardiopulmonary bypass (CPB) technology. These injuries arise from a complex interplay of embolic load, hypoperfusion, impaired cerebral autoregulation, inflammatory responses, and patient-specific vulnerabilities, underscoring the need for a precision-based approach rather than generalized neuroprotective strategies. This narrative review synthesizes current evidence across the perioperative continuum to propose an integrated, data-driven framework for precision neuroprotection. This structured narrative review was conducted according to the SANRA framework. Literature was retrieved from PubMed and Scopus databases covering publications from January 2005 to October 2025.Preoperative risk assessment is enhanced by multimodal imaging, frailty evaluation, neurocognitive screening, metabolic and inflammatory biomarkers, and emerging machine learning models capable of predicting postoperative delirium and cognitive dysfunction. Intraoperatively, precision neuroprotection centers on autoregulation-guided blood pressure management, individualized perfusion targets during CPB, optimized temperature and glucose control, and multimodal neuromonitoring, including transcranial Doppler (TCD), near-infrared spectroscopy (NIRS), processed EEG/BIS, and embolic load characterization, to identify and mitigate cerebral insults in real time. Postoperative precision strategies incorporate biomarker-based surveillance, delirium screening, early mobilization, cognitive rehabilitation, and digital monitoring through wearable devices and remote platforms to detect delayed neurological decline. Although these modalities demonstrate clear potential, major research gaps persist, including variability in device accuracy, limited ability to differentiate embolic subtypes, inconsistent biomarker performance and insufficient integration of AI-driven analytics with clinical workflows. Collectively, available evidence supports the transition from fragmented, organ-focused neuroprotection to a unified, multimodal continuum that leverages data integration, personalized thresholds, and predictive analytics. Future large-scale, human-centered clinical trials are essential to validate this approach and operationalize precision neuroprotection across cardiac surgical practice.