Changes in Neonatal Intraoperative Electroencephalogram Alpha: Delta Ratios Precede Neurologic Injury

Abstract

Background: Unrecognized intraoperative cerebral ischemia during neonatal aortic arch reconstruction may precede neurologic injury. Electroencephalogram (EEG) alpha:delta ratio (A:D) changes predict cerebral ischemia; however, if A:D differences can identify ischemia during neonatal antegrade cerebral perfusion (ACP) and aortic arch reconstruction is unknown. We hypothesized that A:D changes would precede neurologic injury. Methods: Simultaneous EEG derived left versus right: hemispheric and anterior cerebral A:Ds were retrospectively measured at baseline and every 5 min during arterial cannulation, cooling, ACP, and the rewarming phases of the operation. A paired left versus right A:D difference >25% was considered significant for ischemia, and the duration of a significant and continuous A:D difference was quantified in minutes. Neonates were divided into two groups: (1) new neurologic injury (stroke or seizure) and (2) no known neurologic injury. Results: From 72 neonates, there were no significant differences in the baseline A:Ds. Seven neonates (9.7%) developed a new neurologic injury (seizure = 3, stroke = 2, seizure and stroke = 2). Male gender and longer ACP times were significantly associated with neurologic injury. In neonates with a neurologic injury, the duration of a significant and continuous A:D difference was longer within the hemispheric and anterior regions. Multivariable analysis demonstrated that a significant and continuous anterior A:D difference (odds ratio: 1.345, 95% CI 1.058-1.712; P = .01) was independently associated with neurologic injury. Conclusions: A longer continuous anterior A:D difference > 25% was independently associated with neurologic injury. Intraoperative EEG monitoring could be considered during neonatal arch reconstruction.

Keywords

cerebral protection neurologic injury neonates perfusion

Introduction

Neurologic injury following aortic arch reconstruction occurs in 10% to 70% of neonates, often presenting as seizure activity or stroke.^1–6 Despite intraoperative hemodynamic and neurologic monitoring, a postoperative seizure or stroke may be the result of unrecognized cerebral ischemia during the operation. Although an electroencephalogram (EEG) can identify cerebral ischemia within minutes,^7,8 the inclusion of intraoperative EEG monitoring during pediatric cardiac surgery has been limited. Hypothermia-induced cortical suppression during the critical phases of the procedure, in conjunction with the resources required to examine the EEG recordings in real time, have minimized clinical applications.

Quantitative EEG analysis uses Fast Fourier Transform (FFT) to examine EEG signals producing real-time graphical outputs. Although qEEG analysis may have potential superiority over traditional methods of review, signal artifact can be mistakenly incorporated into the analysis and suggest abnormal activity.^9–11 Despite this limitation, quantitative EEG monitoring might be helpful when examining the low voltage alpha and delta frequencies and the corresponding alpha:delta ratio (A:D) that is sensitive and specific for cerebral ischemia.^12,13 Changes in the A:D suggestive of cerebral ischemia¹⁴ correlate with infarct size¹⁵ and warrant urgent intervention to prevent further neurologic injury.¹⁶ However, it is unknown if A:D changes predict neurologic injury in neonates requiring aortic arch reconstruction.

Using quantitative EEG analysis, we retrospectively examined intraoperative EEG recordings from neonates requiring aortic arch reconstruction to identify whether changes in the A:D preceded neurologic injury. We hypothesized that an interhemispheric or regional A:D difference of >25% was significant for ischemia and that longer continuous durations of a significant A:D difference would be associated with neurologic injury.

Patients and Methods

Following Institutional Review Board approval and waiver of consent for this retrospective review, neonates (<30 days) requiring aortic arch reconstruction at the University of Rochester Medical Center using antegrade cerebral perfusion (ACP) between 2015 and 2021 were identified. Neonates that developed a postoperative cardiac arrest or required extracorporeal membrane oxygenation were excluded.¹⁷

Preoperative Neurologic Exam

A baseline exam assessed for gross neurologic deficits at the time of admission. Per the local standard of care, preoperative head ultrasounds were evaluated for abnormalities including, interventricular/intracerebral hemorrhage or mass, and subsequent MRI imaging was obtained when necessary to quantify structural changes. A 30-min baseline EEG, 24 to 48 h prior to surgery, was performed to examine for seizure activity as part of the routine clinical preoperative assessment.

Operative Methods

The operative methods have been previously described^18,19 and are provided in detail within the data supplement.

Electroencephalogram Monitoring

Gold-plated scalp electrodes were fixed using collodion approximately 24 to 48 h before surgery for the baseline study and applied according to the international 10–20 system using the standard temporal, parasagittal, and midline placements except FP1 and FP2.^20–22 Electroencephalogram electrodes were maintained in place for pre-, intra-, and postoperative monitoring. Electroencephalogram recordings were synchronized to the intraoperative events, during four separate phases of the operation: (1) Innominate artery cannulation (Cann); (2) Initiation of cardiopulmonary bypass and cooling (Cool); (3) ACP; and (4) the return to full flow, rewarming, and cessation of cardiopulmonary bypass (Re-warm).

Signal Processing and A:D Calculation

Electroencephalograms were processed using Persyst 14™ software which separates the EEG signal into the faster alpha, and slower delta frequencies (Figure S1). The alpha frequencies are representative of faster electrical activity that is typical during periods of wakefulness. In contrast, the delta frequencies represent slower electrical activity commonly observed during sleep, sedation, or encephalopathy. The A:D is calculated by dividing the quantity of alpha frequency activity by the amount of delta frequency activity. During cerebral ischemia, there is slowing of the EEG waveforms that correspond with a decrease in the alpha frequency, an increase in the delta frequency, and a resulting lower A:D. The A:D is continuously calculated and represents a composite of a running 2-min average and presented as a graphical output using the three predetermined hemispheric, anterior, and posterior regions (Figure 1A-C). Further details regarding how the EEG signal was processed are provided within the data supplement.

A representative EEG during the cooling portion of cardiopulmonary bypass demonstrates the typical increase in the cortical interburst interval (Figure S2), eventually resulting in the absence of cortical bursts and visual isoelectricity (Figure 2A). Despite the absence of cortical bursts, there remains a continued low voltage background frequency that the Persyst software can use to measure the A:D that can be observed visually after magnification of the EEG in Figure 2A. The background frequency (Figure 2B) can be divided into Alpha, Beta, Theta, and Delta components (Figure 2C). From this recording, FFT demonstrates a significantly greater delta component and a reduced alpha component (EEG slowing) from a representative left-sided channel. In comparison, the corresponding right-sided channel has a lower delta and greater alpha component, resulting in a 65%, 137%, and 67% A:D difference within the hemispheric, anterior, and posterior regions (Figure 2D).

Baseline A:D measurements were obtained from the average of three measurements 5 min apart. Although the A:D is provided as a continuous measurement using the Persyst software, the left and right A:Ds were manually recorded every 5 min, and the percent difference calculated. A percent difference > 25% lasting longer than 5 min was considered significant.^14–16 Changes in the A:D < 5 min were thought to reflect periodic changes in cardiopulmonary bypass flow, not considered clinically significant, and therefore not counted. The duration of a significant A:D percent difference for each region (Hemispheric, Anterior, and Posterior) was calculated for each phase of the operation, and the sum for all regional noncontinuous durations were used to generate the overall total A:D difference. Lastly, the longest continuous A:D difference from each region was quantified.

Postoperative Monitoring

Electroencephalogram recordings continued during the postoperative period for 72 h to observe for subclinical seizure activity. Seizures were defined as a rhythmic pattern in one or more leads lasting at least 10 s. All EEG waveforms were reviewed by a neurologist for seizure activity. Neonates were maintained on full cardiopulmonary monitoring with hourly neurologic assessments during the postoperative period. Any concern for focal or abnormal neurologic findings on examination would lead to appropriate neurology consultation and imaging (computed tomography (CT) scan and/or magnetic resonance imaging (MRI) brain imaging).

Statistics

The statistical methods can be found within the data supplement.

Results

Of 72 neonates,7 (9.7%) developed a neurologic injury (subclinical seizure = 3, stroke = 2, subclinical seizure and stroke = 2) (Table S1). Seizure activity was identified in three neonates during the postoperative period, and in two neonates during the operation following the cessation of cardiopulmonary bypass.

Perioperative demographics demonstrated that male gender and longer ACP times were associated with the development of a known neurologic injury (Table 1). One child in each group had a choroid plexus cyst identified during preoperative head ultrasound. There were no significant differences between groups in the ACP flow rate, temperature, or rewarming rate.

Table 1.

Perioperative Demographics.^a

	New neuro in (n = 7)	No known neuro in (n = 65)	P value
Age (days)	9.0 (3.0, 12.0)	9.0 (3.0, 13.0)	.35
Male Gender	7 (100%)	37 (56.9%)	.04
Weight (kg)	3.5 (2.9, 3.8)	3.4 (3.1, 3.6)	.76
Height (cm)	51.0 (48.0, 55.0)	50.0 (48.0, 52.0)	.56
Head circumference (cm)	34.0 (33.5, 35.5)	34.0 (33.3, 35.0)	.54
Gestation (weeks)	39.0 (38.0, 39.0)	39.0 (38.0, 39.0)	.62
Prematurity	0	5, (7.7%)	.99
Extracardiac Abn	2 (28.6%)	17 (26.2%)	.66
Chromosomal Abn	2 (28.6%)	18 (27.7%)	.99
Syndromic	3 (42.9%)	21 (32.3%)	.68
Abn Head US	1 (14.3%)	1 (1.5%)	.19
Primary Procedure
Norwood	4 (57.1%)	22 (33.9%)	.24
AAH +/- VSD	2 (28.6%)	26 (40.0%)	.70
IAA	1 (14.3%)	6 (9.2%)	.99
other	0	11 (16.9%)	.58
Single Ventricle	4 (57.1%)	29 (44.6%)	.69
Operative Details
CPB time (min)	168.0 (146.0, 192.0)	128.0 (105.5, 187.0)	.16
X-clamp time (min)	48.0 (46.0, 95.0)	55.0 (37.5, 77.0)	.57
ACP time (min)	83.0 (40.0, 93.0)	31.0 (27.5, 74.0)	.04
ACP flow (cc/min)	30 (30, 35)	30.0(25.0, 30.0)	.15
ACP temp °C	22.0 (22.0, 24.0)	22.0 (22.0, 24.0)	.96
Nadir nasal temp °C	22.1(21.5, 24.0)	22.1 (21.5, 24.0)	.81
Nadir core temp °C	24.4 (22.8, 25.0)	24.4 (23.3, 25.7)	.58
Cooling time (min)	29.0 (20.0,30.0)	22.0 (19.5, 27.5)	.23
Rewarming time (min)	40.0 (35.0, 50.0)	36.0 (30.0, 39.0)	.47

Abbreviations:AAH, aortic arch hypoplasia; Abn, abnormality;ACP, antegrade cerebral perfusion; CPB, cardiopulmonary bypass; IAA, interrupted aortic arch; neuro in, neurologic injury; US, ultrasound; VSD, ventricular septal defect; X-clamp, aortic cross-clamp.

Data presented as median with interquartile range or frequency and percentage.

Electroencephalograms

Baseline EEGs were similar between groups; no patient in either group had a baseline A:D difference of >25% or evidence of seizure activity (Table S2). A representative example of the evolution of the A:D during the study period from an infant whose postoperative EEG demonstrated persistent asynchrony and who was diagnosed with a parietal stroke by CT scan is provided in Figure 3A. During the cooling period of cardiopulmonary bypass, there was a 35-min continuous anterior A:D difference and a 25-min posterior and hemispheric A:D difference > 25. In sharp contrast, Figure 3B demonstrates the hemispheric, anterior, and posterior A:Ds of a neonate who did not develop a known neurologic injury. From this neonate, although the A:D fluctuated in all leads over the study period, there were no continuous or significant periods with a 25% A:D difference.

Timing of A:D Difference

From the 72 neonates, 35 (48.6%) did not demonstrate a significant A:D difference. From the 37 neonates (51.4%) that did develop a significant A:D difference, the majority were composed of short (5-10 min) noncontinuous intervals over the study period. A significant A:D was present in at least one neonate during all phases of the operation except the baseline (Figure S3). Although the mean duration of a significant A:D difference was not statistically significant during arterial cannulation, 3 (4.2%) of neonates had a 10- to 20-min significant and continuous A:D difference during this phase. All three developed neurologic injuries, and in one neonate, the significant A:D difference continued into the cooling period. Variations between the hemispheric, anterior, and posterior A:D were observed without consistent uniformity between the cooling and ACP phases. However, during the rewarming phase, the total duration of an A:D difference was significantly greater.

Alpha:Delta Difference and Neurologic Injury

The percentage of neonates who developed an anterior A:D difference was significantly greater in those who developed neurologic injury (Figure 4A). Further, both the total duration of significant A:D differences and the duration of a continuous A:D difference were significantly greater in the hemispheric and anterior regions of neonates who developed neurologic injury (Figure 4B and C, Table S3). However, there was no significant difference comparing the total or continuous durations of an A:D difference within the posterior region. The cumulative totals from all three regions and the cumulative total of a continuous A:D difference were also significantly greater within those who developed a neurologic injury (Figure 4D). The impact of combining the data from the different regions on the outcome of neurologic injury is provided within the data supplement (Figure S4A and S4B).

To understand if significant A:D differences from the hemispheric, anterior, and posterior regions predict neurologic injury, receiver operating charateristic curves were created (Figure 5A and 5B). The total A:D difference duration from the hemispheric and posterior regions were not predicative. However, the total duration of a significant anterior A:D difference was predictive in identifying neurologic injury (AUC = 0.802; P = .009). Similarly, although the duration of the longest continuous A:D difference within the hemispheric and posterior regions was not significant, a continuous A:D difference from within the anterior region was highly predictive for neurologic injury (AUC = 0.825; P = .005). When using a threshold of 12.5 continuous minutes of an A:D anterior difference, the sensitivity and specificity of predicting a neurologic injury was 71.4% and 4.6%, respectively. Data regarding the impact of predicting neurologic injury using the A:D difference from the combination of two or more regions are provided within the data supplement (Table S4, Figures S5A and S5B). Multivariable analysis demonstrated that only a longer continuous anterior A:D difference was independently associated with neurologic injury (Table 2).

Table 2.

Multivariate Analysis.

Variable	Odds ratio	95% confidence interval	P value
Age	0.811	0.484, 1.360	.427
Weight	0.038	0.000, 19.013	.303
Gender	—	—	.998
CPB	1.006	0.980, 1.033	.665
ACP	1.047	0.985, 1.113	.142
Long hemi	1.042	0.843, 1.287	.706
Long ant	1.345	1.058, 1.712	.016
Long post	1.003	0.778, 1.293	.981

Abbreviations: ACP, antegrade cerebral perfusion; ant, anterior; CPB, cardiopulmonary bypass; hemi, hemispheric; post, posterior.

Comment

The findings from this manuscript suggest that a continuous anterior EEG A:D difference of greater than 25% for more than 12.5 min is predictive and independently associated with neurologic injury. Although postoperative EEG monitoring to identify subclinical seizure activity has become more common,^5,6 intraoperative EEG monitoring has been used sparingly.²⁰ Longer intraoperative isoelectric periods, when cortical activity is visually absent, have been linked to both postoperative neurologic injury¹ and neurodevelopmental delay.²¹ When present prior to ACP, a nonisoelectric EEG may be associated with white matter injury.²² Although historically^20,21 and within this cohort, we have observed isoelectricity prior to ACP at approximately 25 °C, the timing of isoelectricity has varied in both neonatal and animal data^22,23 and may be related to EEG amplitude thresholds, magnification, and/or rates of cooling. In addition, a greater percentage of intraoperative cortical asynchronous bursts, particularly during rewarming, have been associated with postoperative seizures²⁰ and neurologic dysfunction.²⁴ Collectively, applying these parameters to clinical practice requires a dedicated neurologist to review the study in real time, limiting the practicality of implementing these measures. Although the prior use of quantitative EEG monitoring has provided more detailed information as well as graphical outputs, these measures often required file processing after completion of the EEG study, limiting the clinical applicability.^14,25–27 Most important is that historically, the hypothermia-induced cortical suppression limited EEG information during the critical cooling, ACP, and rewarming portions of the surgery.

In sharp contrast, although these data were analyzed retrospectively, comparing the A:D using the Persyst software in real time provides easily interpretable graphical outputs that can be assessed by an anesthesiologist, perfusionist, or surgeon. This could allow for timely intervention during critical phases of the operation, prompting changes in the cardiopulmonary bypass flow, temperature, and CO₂. During cerebral ischemia, the faster and often dominant alpha components of the background frequency decrease in conjunction with an increase in the delta components. Subsequently, changes to the A:D have been closely related to cerebral ischemia in animal models,²⁸ case reports,²⁹ and clinical series.^1,14,30 These changes can be observed in as early as 30 to 60 s,¹⁴ correlate with cerebral infarct size,¹⁵ and prompt intervention can normalize the A:D. Other modalities of intraoperative neuromonitoring, include cerebral near-infrared spectroscopy (NIRS) and transcranial Doppler that have also shown benefit in predicting neurologic injury by measuring cerebral oxygen saturation and the blood flow velocity from within the anterior cerebral circulation,^31–33 where interestingly, the greatest A:D differences also occurred. Although not examined within this manuscript, the use of quantitative EEG monitoring in conjunction with NIRS and transcranial Doppler may be helpful in confirming neurologic injury.

Interestingly, two neonates did not develop overt A:D changes during surgery but developed neurologic injury. One patient developed seizures during the postoperative period, originating centrally within the brain. This area may represent a “blind spot” when comparing left versus right A:D differences and warrants further investigation. Alternatively, a global left and right cerebral hemispheric ischemic event may not have been identified when calculating the A:D difference. A second patient developed an intraoperative seizure after two separate and prolonged periods of ACP. Therefore, the mechanism(s) of the seizure and corresponding bilateral microhemorrhages in this patient may be unrelated to ischemia, but rather edema. It is interesting that changes in the posterior A:D were not predictive of significant neurologic events. Further investigation and screening MRIs are necessary to more thoroughly evaluate the impact of isolated posterior A:D changes.

When comparing the baseline A:D prior to surgery, there were no significant differences between neonates who developed postoperative neurologic injury. Prior studies have suggested that changes, particularly in the EEG background frequency, are associated with poor postoperative outcomes.¹ However, background frequency changes may be suppressed by sedation,^24,27 and therefore, these changes may reflect the overall acuity of urgent surgical cases within the hospital and not distinct EEG features.

Despite the theoretical benefit of ACP over deep hypothermic circulatory arrest, no conclusive evidence demonstrates its superiority.³ This disparity may reflect the need to occlude the innominate artery during ACP cannulation. Three neonates within this series developed significant A:D changes during the clamping of the innominate artery, which continued in one neonate during the cooling period of cardiopulmonary bypass. Although an intact Circle of Willis should maintain adequate right cerebral blood flow from an unobstructed left carotid artery, preoperative congenital abnormalities in the cerebral artery architecture^34,35 in conjunction with either bidirectional or retrograde aortic arch flow from a patent ductus arteriosus, may limit oxygen delivery to the right anterior cerebrum during ACP cannulation. Follow-up cerebral imaging to determine the anatomy of the Circle of Willis was only available from one neonate following neurologic injury, and in that single patient, the Circle of Willis was patent.

Limitations

There are several limitations to this work. This was a retrospective study and is therefore limited by that approach. Baseline EEG measurements were only 30 min in duration, and longer surveillance may have captured significant differences. Most important, MRI imaging was only obtained when clinically indicated, and therefore, a greater number of subclinical neurologic events could have occurred either pre- or postoperatively. Electroencephalogram differences were not linked to other neuromonitoring modalities such as cerebral oximetry or transcranial Doppler. Our standard practice over the study period was to utilize left NIRS measurements; and therefore, a percent difference comparison between NIRS and A:D measurements was not possible. Last, the intraoperative EEG recordings were analyzed in greater detail when compared with the postoperative EEG recordings resulting in a bias for the prediction of neurologic injury during the operative period.

Conclusion

Significant EEG A:D differences suggestive of ischemia were observed during all aspects of the surgical procedure, including the temporary occlusion of the innominate artery for arterial cannulation. In addition, a longer continuous left versus right anterior A:D difference was highly predictive and independently associated with the development of neurologic injury.

Quantitative EEG monitoring may be a sensitive tool to detect cerebral ischemia, optimize operative management, identify neonates requiring heightened postoperative surveillance, and could be considered in conjunction with other intraoperative neuromonitoring modalities.

Supplemental Material

sj-docx-1-pch-10.1177_21501351241269963 - Supplemental material for Changes in Neonatal Intraoperative Electroencephalogram Alpha: Delta Ratios Precede Neurologic Injury

Supplemental material, sj-docx-1-pch-10.1177_21501351241269963 for Changes in Neonatal Intraoperative Electroencephalogram Alpha: Delta Ratios Precede Neurologic Injury by Michael F. Swartz, Justin Lansinger, Emelie-Jo Scheffler, Aubrey Duncan, Jill M. Cholette, Shuichi Yoshitake and George M. Alfieris in World Journal for Pediatric and Congenital Heart Surgery

Footnotes

Abbreviations

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Michael F. Swartz

Emelie-Jo Scheffler

Supplemental Material

Supplemental material for this article is available online.

References

Lin

Feng

, et al. Perioperative EEG background and discharge abnormalities in children undergoing cardiac surgery: a prospective single-centre observational study. Br J Anaesth. 2023;131(2):360-372.

Alkhatip

AAAMM

Kamel

Farag

, et al. Deep hypothermic circulatory arrest in the pediatric population undergoing cardiac surgery with electroencephalography monitoring: a systematic review and meta-analysis. J Cardiothorac Vasc Anesth. 2021;35(10):2875-2888.

Andropoulos

Easley

Gottlieb

Brady

. Neurologic injury in neonates undergoing cardiac surgery. Clin Perinatol. 2019;46(4):657-671.

Algra

Jansen

van der Tweel

, et al. Neurological injury after neonatal cardiac surgery: a randomized, controlled trial of 2 perfusion techniques. Circulation. 2014;129(2):224-233.

Naim

Gaynor

Chen

, et al. Subclinical seizures identified by postoperative electroencephalographic monitoring are common after neonatal cardiac surgery. J Thorac Cardiovasc Surg. 2015;150(1):169-178. discussion 178–80.

Naim

Putt

Abend

, et al. Development and validation of a seizure prediction model in neonates after cardiac surgery. Ann Thorac Surg. 2021;111(6):2041-2048.

Blume

Ferguson

McNeill

. Significance of EEG changes at carotid endarterectomy. Stroke. 1986;17(5):891-897.

Domenick Sridharan

Thirumala

Chaer

, et al. Predictors of cross-clamp-induced intraoperative monitoring changes during carotid endarterectomy using both electroencephalography and somatosensory evoked potentials. J Vasc Surg. 2018;67(1):191-198.

Sharpe

Davis

Reiner

, et al. Assessing the feasibility of providing a real-time response to seizures detected with continuous long-term neonatal electroencephalography monitoring. J Clin Neurophysiol. 2019;36(1):9-13.

10.

Scheuer

Wilson

Antony

Ghearing

Urban

Bagić

. Seizure detection: interreader agreement and detection algorithm assessments using a large dataset. J Clin Neurophysiol. 2021;38(5):439-447.

11.

Sierra-Marcos

Scheuer

Rossetti

. Seizure detection with automated EEG analysis: a validation study focusing on periodic patterns. Clin Neurophysiol. 2015;126(3):456-462.

12.

Finnigan

Wong

Read

. Defining abnormal slow EEG activity in acute ischaemic stroke: delta/alpha ratio as an optimal QEEG index. Clin Neurophysiol. 2016;127(2):1452-1459.

13.

Khan

Dar

Johnson

, et al. Correlations between quantitative EEG parameters and cortical blood flow in patients undergoing extracorporeal membrane oxygenation with and without encephalopathy. J Clin Neurophysiol. 2023. Epub ahead of print. PMID: 37934074. doi. 10.1097/WNP.0000000000001035

14.

Kamitaki

Wong

Mendiratta

Choi

. Quantitative EEG changes correlate with post-clamp ischemia during carotid endarterectomy. J Clin Neurophysiol. 2021;38(3):213-220.

15.

Shreve

Kaur

, et al. Electroencephalography measures are useful for identifying large acute ischemic stroke in the emergency department. J Stroke Cerebrovasc Dis. 2019;28(8):2280-2286.

16.

Dickey

Mitsias

Olango

, et al. The prognostic value of quantitative EEG in patients undergoing mechanical thrombectomy for acute ischemic stroke. J Clin Neurophysiol. 2022;39(4):276-282.

17.

Polimenakos

Wojtyla

Smith

, et al. Post-cardiotomy extracorporeal cardiopulmonary resuscitation in neonates with complex single ventricle: analysis of outcomes. Eur J Cardiothorac Surg. 2011;40(6):1396-1405. discussion 1405.

18.

Swartz

Atallah-Yunes

Meagher

, et al. Surgical strategy for aortic coarctation repair resulting in physiologic arm and leg blood pressures. Congenit Heart Dis. 2011;6(6):583-591.

19.

Walters

Swartz

Patel

Cholette

Atallah-Yunes

Alfieris

. Risk factors for a positive neoaortic arch gradient after stage I palliation for hypoplastic left heart syndrome. Ann Thorac Surg. 2013;96(2):645-650.

20.

Seltzer

Swartz

Kwon

Burchfiel

Alfieris

Guillet

. Intraoperative electroencephalography predicts postoperative seizures in infants with congenital heart disease. Pediatr Neurol. 2014;50(4):313-317.

21.

Seltzer

Swartz

Kwon

, et al. Neurodevelopmental outcomes after neonatal cardiac surgery: role of cortical isoelectric activity. J Thorac Cardiovasc Surg. 2016;151(4):1137-1142.

22.

Laurent

Mensah-Brown

, et al. Electroencephalography as a tool to predict cerebral oxygen metabolism during deep-hypothermic circulatory arrest in neonates with critical congenital heart disease. JTCVS Open. 2023;16:801-809.

23.

Mavroudis

Mensah-Brown

, et al. Electroencephalographic response to deep hypothermic circulatory arrest in neonatal swine and humans. Ann Thorac Surg. 2018;106(6):1841-1846.

24.

Swartz

Seltzer

Cholette

, et al. Intraoperative cortical asynchrony predicts abnormal postoperative electroencephalogram. Ann Thorac Surg. 2021;111(2):645-654.

25.

McDevitt

Gul

Jones

Scholefield

Seri

Drury

. Perioperative electroencephalography in cardiac surgery with hypothermic circulatory arrest: a narrative review. Interact Cardiovasc Thorac Surg. 2022;35(4):ivac198.

26.

Tian

Liu

Zhou

, et al. The change index of quantitative electroencephalography for evaluating the prognosis of large hemispheric infarction. J Integr Neurosci. 2021;20(2):341-347.

27.

Chao

Gutiérrez

Legatt

, et al. Decreased electroencephalographic alpha power during anesthesia induction is associated with EEG discontinuity in human infants. Anesth Analg. 2022;135(6):1207-1216.

28.

Ferreira

Mattos

Jóia de Mello

, et al. Increased relative delta bandpower and delta indices revealed by continuous qEEG monitoring in a rat model of ischemia-reperfusion. Front Neurol. 2021;12:645138.

29.

Hussein

. Relative alpha variability changes precede alpha-delta ratio changes in cerebral ischemia. J Stroke Cerebrovasc Dis. 2020;29(11):105262.

30.

Zhang

Chen

Xie

, et al. Application of quantitative EEG in acute ischemic stroke patients who underwent thrombectomy: a comparison with CT perfusion. Clin Neurophysiol. 2022;141:24-33.

31.

Hibino

Peterson

Tachibana

, et al. Association of cerebral oximetry with brain ischemic lesions and functional outcomes in arch repair. Ann Thorac Surg. 2024;117(3):627-633.

32.

De Silvestro

Krüger

Steger

, et al. Cerebral desaturation during neonatal congenital heart surgery is associated with perioperative brain structure alterations but not with neurodevelopmental outcome at 1 year. Eur J Cardiothorac Surg. 2022;62(5):ezac 138.

33.

Rilinger

Smith

deRegnier

RAO

, et al. Transcranial Doppler identification of neurologic injury during pediatric extracorporeal membrane oxygenation therapy. J Stroke Cerebrovasc Dis. 2017;26(10):2336-2345.

34.

Zaninovich

Ramey

Walter

Dumont

. Completion of the Circle of Willis varies by gender, age, and indication for computed tomography angiography. World Neurosurg. 2017;106:953-963.

35.

Sonobe

Kato

Mathis

Hiramatsu

. Surgical correction of coarctation of the aorta featuring an incomplete Circle of Willis. Interact Cardiovasc Thorac Surg. 2020;30(5):794-795.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

4.13 MB