Near-Infrared Spectroscopy Used During Cardiopulmonary Resuscitation: Instrumentation, Signal Metrics, Clinical Context, and Feasibility: A Scoping Review

Conventional cardiopulmonary resuscitation (CPR) is guided primarily by process metrics that do not directly quantify cerebral hemodynamics or perfusion. Near-infrared spectroscopy (NIRS) provides continuous, non-invasive monitoring of regional tissue oxygenation and has emerged as a candidate modality for physiologic feedback during low-flow states. However, CPR applications vary across devices and signal processing. This scoping review maps how NIRS has been implemented during conventional CPR in humans and porcine models, with emphasis on instrumentation characteristics, signal processing, acquisition bandwidth, artifact handling, physiologic associations, and feasibility constraints. From 1048 records, 39 studies met the inclusion criteria. Most used forehead-based cerebral rSO2 monitoring (30/39). Rising cerebral oxygenation trajectories were consistently associated with return of spontaneous circulation (ROSC). In contrast, persistently low or non-increasing patterns were associated with non-ROSC, and absolute thresholds varied substantially across devices and studies. A minority of investigations derived compression-rate or waveform features from hemoglobin signals. Feasibility findings emphasized rapid probe placement without interrupting compressions but highlighted motion artifact, workflow constraints, and incomplete acquisition reporting. Overall, during conventional CPR, NIRS primarily serves as a dynamic monitor of oxygenation trends rather than a validated prognostic tool. Emerging waveform-based and hemodynamic analyses suggest the potential to evaluate CPR efficiency using perfusion-responsive optical features.