Variane et al. comprehensively describe the current uses of neuromonitoring in Neuro-NICUs focusing on aEEG, cEEG, and NIRS. Multimodality monitoring simultaneously with aEEG or cEEG and NIRS is likely to improve understanding of the physiology of both functional and hemodynamic changes and the resulting risk of cerebral injury. The incorporation of other time-synchronized physiologic vital signs will enhance the development of neuroprotective strategies. The available options, as well as challenges of data integration and processing, are described and the power of artificial intelligence and machine learning to significantly benefit vulnerable critically ill neonates is highlighted. The authors conclude that the broader use of brain monitoring with analysis of large amounts of clinical data available in an intensive care setting has the potential to significantly change care and improve outcomes.

Given the significant risk for adverse outcomes in newborns with HIE, including death and neurodevelopmental impairment, a wide range of neurodiagnostic modalities have been evaluated to provide critical diagnostic and prognostic information on brain injury and prediction of outcome. Chock et al. provide a comprehensive review of continuous measures including cEEG, aEEG, NIRS, and heart rate variability as well as serial measures including cranial ultrasound and somatic and visual evoked potentials in newborns with HIE undergoing therapeutic hypothermia. cEEG and aEEG are promising predictors of adverse outcome with high specificity and sensitivity. The higher specificity seen with cEEG is offset by its greater complexity and cost for placement and interpretation. Mutli-modality use of both cEEG or aEEG with NIRS along with physiologic vital signs is appealing as it allows for continuous bedside assessment of both cerebral function and cerebral oxygenation and correlates them with alterations in other physiologic parameters.

NIRS monitoring has been investigated in premature infants in different situations including the transitional period after birth (1–4), for diagnosis of a hemodynamically significant patent ductus arteriosus (5), or for intraventricular hemorrhage or post-hemorrhagic hydrocephalus (6, 7). Several observational studies have also demonstrated an association between early cerebral hypoxia and adverse outcomes including neurodevelopment (4, 8–10). While the SafeBoosC randomized clinical trial (RCT) demonstrated a reduction in the burden of cerebral hypoxia using a NIRS interventional guideline to maintain targeted cerebral saturation values (1), current evidence from large, multinational RCTs does not demonstrate a reduction in adverse outcomes including mortality or brain injury (2, 3, 11). Greisen et al. provide a perspective on the state of NIRS monitoring for the preterm infant. The authors compare cerebral oximetry to other monitoring modalities that may be used in the neonatal intensive care setting despite minimal empiric evidence of clinical benefit, including pulse oximetry, non-invasive electric cardiometry, and invasive blood pressure monitoring. Future research may require a focus on more granular effects of a targeted cerebral NIRS algorithm on the brain and clarify timing and conditions for which NIRS monitoring may best optimize care for the premature infant.

Neuromonitoring with NIRS may further have utility in infants with CHD, particularly in the peri-operative period and may reduce the need for intubations in select infants with CHD (12, 13). Cerebral hypoxia has been associated with adverse neuroimaging outcomes and impaired neurodevelopment in the CHD population (14–16). Hoffman et al. describe the utility of both cerebral and somatic NIRS monitoring with relation to carbon dioxide tension in 178 infants with hypoplastic left heart syndrome following Norwood palliation. Increasing arterial carbon dioxide tension was associated with increased cerebral but decreased renal oxygenation. Along with this differential response to carbon dioxide, the magnitude of these responses was greater in the early 18 h post-operative period. Given the sensitivity of the brain to hypocarbia with resulting cerebral vasoconstriction, strict attention to carbon dioxide levels may be an important strategy to optimize cerebral blood flow. The authors speculate that normalization of cerebral hemodynamics may be achieved in the early post-operative period with dynamic manipulation of carbon dioxide levels through directed ventilator approaches such as permissive hypercapnia.

Continuous neuromonitoring using NIRS and aEEG or cEEG allows for bedside assessment of brain, renal, and splanchnic/mesenteric oxygenation and perfusion as well as cerebral function and seizure detection. Further combination with vital sign monitoring including blood pressure, pulse oximetry, heart rate and temperature is described as multimodal monitoring and provides further understanding of physiology. These approaches have been previously described in premature infants and in newborns with HIE (4, 17, 18). Variane et al. describe 10 additional cases including cardiopulmonary disorders, circulatory disorders, and abdominal disorders, where multimodal monitoring with multi-site NIRS and in some cases with aEEG provided earlier recognition of the underlying hemodynamic status and its impact.

Impairment of cerebral autoregulation may lead to alterations in cerebral blood flow, particularly in vulnerable populations including neonates with HIE, preterm infants, and those with congenital heart disease. Leon et al. review various modalities to assess cerebral blood flow and specifically cerebral autoregulatory capacity in these conditions. Doppler ultrasound, magnetic resonance imaging, and NIRS are techniques that have advantages and limitations for use in the fetal and neonatal period. The authors emphasize that future assessment of cerebral autoregulation will require an individualized approach and use of multiple synchronized monitoring modalities and predictive analytics.

Several uses for cEEG in the neonatal population have been described, including seizure diagnosis and management, identifying neonates at risk for seizures, and prognostication by assessment of background activity. It is the standard of care for seizure diagnosis as designated by the World Health Organization (WHO), International League Against Epilepsy (ILAE), and American Clinical Neurophysiology Society (ACNS) (19–21). Sandoval et al. acknowledge that high cost is a factor limiting more widespread use of cEEG. In comparison, aEEG is recognized as often helpful when used together with cEEG, as it allows bedside providers to review the simplified tracing without the need for neurophysiologists. However, short, low amplitude seizures or seizures in areas of the brain not monitored by aEEG will be missed, and a high skill level with aEEG interpretation is required. Future uses of cEEG include centralized cEEG interpretation, automated seizure detection, and prenatal EEG use.