Quantitative analysis of high-frequency activity in neonatal EEG.

OBJECTIVE: To determine the presence and potential utility of independent high-frequency activity recorded from scalp electrodes in the electroencephalogram (EEG) of newborns. METHODS: We compare interburst intervals and continuous activity at different frequencies for EEGs retrospectively recorded at 256 Hz from 4 newborn groups: 1) 36 preterms (<32 weeks' gestational age, GA); 2) 12 preterms (32-37 weeks' GA); 3) 91 healthy full terms; 4) 15 full terms with hypoxic-ischemic encephalopathy (HIE). At 4 standard frequency bands (delta, 0.5-3 Hz; theta, 3-8 Hz; alpha, 8-15 Hz; beta, 15-30 Hz) and 3 higher-frequency bands (gamma1, 30-48 Hz; gamma2, 52-99 Hz; gamma3, 107-127 Hz), we compared power spectral densities (PSDs), quantitative features, and machine learning model performance. Feature selection and further machine learning methods were performed on one cohort. RESULTS: We found significant (P < 0.01) differences in PSDs, quantitative analysis, and machine learning modelling at the higher-frequency bands. Machine learning models using only high-frequency features performed best in preterm groups 1 and 2 with a median (95% confidence interval, CI) Matthews correlation coefficient (MCC) of 0.71 (0.12-0.88) and 0.66 (0.36-0.76) respectively. Interburst interval-detector models using both high- and standard-bandwidths produced the highest median MCCs in all four groups. High-frequency features were largely independent of standard-bandwidth features, with only 11/84 (13.1%) of correlations statistically significant. Feature selection methods produced 7 to 9 high-frequency features in the top 20 feature set. CONCLUSIONS: This is the first study to identify independent high-frequency activity in newborn EEG using in-depth quantitative analysis. Expanding the EEG bandwidths of analysis has the potential to improve both quantitative and machine-learning analysis, particularly in preterm EEG.

Machine Learning Detects Intraventricular Haemorrhage in Extremely Preterm Infants.

OBJECTIVE: To test the potential utility of applying machine learning methods to regional cerebral (rcSO2) and peripheral oxygen saturation (SpO2) signals to detect brain injury in extremely preterm infants. STUDY DESIGN: A subset of infants enrolled in the Management of Hypotension in Preterm infants (HIP) trial were analysed (n = 46). All eligible infants were <28 weeks' gestational age and had continuous rcSO2 measurements performed over the first 72 h and cranial ultrasounds performed during the first week after birth. SpO2 data were available for 32 infants. The rcSO2 and SpO2 signals were preprocessed, and prolonged relative desaturations (PRDs; data-driven desaturation in the 2-to-15-min range) were extracted. Numerous quantitative features were extracted from the biosignals before and after the exclusion of the PRDs within the signals. PRDs were also evaluated as a stand-alone feature. A machine learning model was used to detect brain injury (intraventricular haemorrhage-IVH grade II-IV) using a leave-one-out cross-validation approach. RESULTS: The area under the receiver operating characteristic curve (AUC) for the PRD rcSO2 was 0.846 (95% CI: 0.720-0.948), outperforming the rcSO2 threshold approach (AUC 0.593 95% CI 0.399-0.775). Neither the clinical model nor any of the SpO2 models were significantly associated with brain injury. CONCLUSION: There was a significant association between the data-driven definition of PRDs in rcSO2 and brain injury. Automated analysis of PRDs of the cerebral NIRS signal in extremely preterm infants may aid in better prediction of IVH compared with a threshold-based approach. Further investigation of the definition of the extracted PRDs and an understanding of the physiology underlying these events are required.

Neonatal EEG graded for severity of background abnormalities in hypoxic-ischaemic encephalopathy.

This report describes a set of neonatal electroencephalogram (EEG) recordings graded according to the severity of abnormalities in the background pattern. The dataset consists of 169 hours of multichannel EEG from 53 neonates recorded in a neonatal intensive care unit. All neonates received a diagnosis of hypoxic-ischaemic encephalopathy (HIE), the most common cause of brain injury in full term infants. For each neonate, multiple 1-hour epochs of good quality EEG were selected and then graded for background abnormalities. The grading system assesses EEG attributes such as amplitude, continuity, sleep-wake cycling, symmetry and synchrony, and abnormal waveforms. Background severity was then categorised into 4 grades: normal or mildly abnormal EEG, moderately abnormal EEG, majorly abnormal EEG, and inactive EEG. The data can be used as a reference set of multi-channel EEG for neonates with HIE, for EEG training purposes, or for developing and evaluating automated grading algorithms.

Parent-led massage and sleep EEG for term-born infants: A randomized controlled parallel-group study.

AIM: To examine the impact of parent-led massage on the sleep electroencephalogram (EEG) features of typically developing term-born infants at 4 months. METHOD: Infants recruited at birth were randomized to intervention (routine parent-led massage) and control groups. Infants had a daytime sleep EEG at 4 months and were assessed using the Griffiths Scales of Child Development, Third Edition at 4 and 18 months. Comparative analysis between groups and subgroup analysis between regularly massaged and never-massaged infants were performed. Groups were compared for sleep stage, sleep spindles, quantitative EEG (primary analysis), and Griffiths using the Mann-Whitney U test. RESULTS: In total, 179 out of 182 infants (intervention: 83 out of 84; control: 96 out of 98) had a normal sleep EEG. Median (interquartile range) sleep duration was 49.8 minutes (39.1-71.4) (n = 156). A complete first sleep cycle was seen in 67 out of 83 (81%) and 72 out of 96 (75%) in the intervention and control groups respectively. Groups did not differ in sleep stage durations, latencies to sleep and to rapid eye movement sleep. Sleep spindle spectral power was greater in the intervention group in main and subgroup analyses. The intervention group showed greater EEG magnitudes, and lower interhemispherical coherence on subgroup analyses. Griffiths assessments at 4 months (n = 179) and 18 months (n = 173) showed no group differences in the main and subgroup analyses. INTERPRETATION: Routine massage is associated with distinct functional brain changes at 4 months. WHAT THIS PAPER ADDS: Routine massage of infants is associated with differences in sleep electroencephalogram biomarkers at 4 months. Massaged infants had higher sleep spindle spectral power, greater sleep EEG magnitudes, and lower interhemispherical coherence. No differences between groups were observed in total nap duration or first cycle macrostructure.

Machine learning for the early prediction of infants with electrographic seizures in neonatal hypoxic-ischemic encephalopathy.

OBJECTIVE: To assess if early clinical and electroencephalography (EEG) features predict later seizure development in infants with hypoxic-ischemic encephalopathy (HIE). METHODS: Clinical and EEG parameters <12 h of birth from infants with HIE across eight European Neonatal Units were used to develop seizure-prediction models. Clinical parameters included intrapartum complications, fetal distress, gestational age, delivery mode, gender, birth weight, Apgar scores, assisted ventilation, cord pH, and blood gases. The earliest EEG hour provided a qualitative analysis (discontinuity, amplitude, asymmetry/asynchrony, sleep-wake cycle [SWC]) and a quantitative analysis (power, discontinuity, spectral distribution, inter-hemispheric connectivity) from full montage and two-channel amplitude-integrated EEG (aEEG). Subgroup analysis, only including infants without anti-seizure medication (ASM) prior to EEG was also performed. Machine-learning (ML) models (random forest and gradient boosting algorithms) were developed to predict infants who would later develop seizures and assessed using Matthews correlation coefficient (MCC) and area under the receiver-operating characteristic curve (AUC). RESULTS: The study included 162 infants with HIE (53 had seizures). Low Apgar, need for ventilation, high lactate, low base excess, absent SWC, low EEG power, and increased EEG discontinuity were associated with seizures. The following predictive models were developed: clinical (MCC 0.368, AUC 0.681), qualitative EEG (MCC 0.467, AUC 0.729), quantitative EEG (MCC 0.473, AUC 0.730), clinical and qualitative EEG (MCC 0.470, AUC 0.721), and clinical and quantitative EEG (MCC 0.513, AUC 0.746). The clinical and qualitative-EEG model significantly outperformed the clinical model alone (MCC 0.470 vs 0.368, p-value .037). The clinical and quantitative-EEG model significantly outperformed the clinical model (MCC 0.513 vs 0.368, p-value .012). The clinical and quantitative-EEG model for infants without ASM (n = 131) had MCC 0.588, AUC 0.832. Performance for quantitative aEEG (n = 159) was MCC 0.381, AUC 0.696 and clinical and quantitative aEEG was MCC 0.384, AUC 0.720. SIGNIFICANCE: Early EEG background analysis combined with readily available clinical data helped predict infants who were at highest risk of seizures, hours before they occur. Automated quantitative-EEG analysis was as good as expert analysis for predicting seizures, supporting the use of automated assessment tools for early evaluation of HIE.

Agreement of Cardiac Output Estimates between Electrical Cardiometry and Transthoracic Echocardiography in Very Preterm Infants.

INTRODUCTION: The aim was to evaluate the agreement between cardiac output estimates obtained by electrical cardiometry (EC) and transthoracic echocardiography (TTE) in very preterm infants. METHODS: This is a single-center prospective observational study in infants born<32 weeks gestational age within 48 h of birth. Continuous EC was recorded and simultaneous TTE obtained on day 1 and day 2 of life. Blinded TTE measurements were performed within a 10 s timeframe using beat-to-beat EC data. The primary outcome was %error of left ventricular (LV) output in milliliters per kilogram per minute (cardiac index (CI)) obtained by TTE compared to LV-CI from EC. Secondary outcome parameters were bias, %bias, limits of agreement and include measures of right ventricular (RV) output and LV systolic time intervals. RESULTS: Analysis was performed for 34 infants (median (IQR) gestational age 29 + 0 (24 + 5 to 30 + 6) weeks + days, birthweight 960 (748 to 1,490) grams) including 44 pairwise LV output measurements on 24 participants (22 on day 1 and day 2). The %error was 54% for LV-CI (EC: 214 (38) mL/kg/min vs. TTE: 163 (47) mL/kg/min). The %error was 78% for RV-CI (EC: 213 (37) mL/kg/min vs. TTE: 241 (77) mL/kg/min). While only LV-CI values affected LV-CI bias, signal quality, heart rate, and RV-CI values affected RV-CI bias. CONCLUSION: EC is not interchangeable with TTE to estimate indices of LV or RV output in very preterm infants within the first 48 h postnatally. EC may not measure LV output distinctly in very preterm infants with intra- and extracardiac shunts.

Evolution of early cerebral NIRS in hypoxic ischaemic encephalopathy.

AIM: To describe early cerebral oxygenation (cSO2 ) and fractional tissue oxygen extraction (FTOE) values and their evolution over the first days of life in infants with all grades of hypoxic-ischaemic encephalopathy (HIE) and to determine whether cSO2 and FTOE measured early (6 and 12 h) can predict short-term outcome. METHODS: Prospective, observational study of cerebral near-infrared spectroscopy (NIRS) in infants >36 weeks' gestation with HIE. Ten one-hour epochs of cSO2 and FTOE were extracted for each infant over the first 84 h. Infants with moderate and severe HIE received therapeutic hypothermia (TH). Abnormal outcome was defined as abnormal magnetic resonance imaging (MRI) and/or death. RESULTS: Fifty-eight infants were included (28 mild, 24 moderate, 6 severe). Median gestational age was 39.9 weeks (IQR 38.1-40.7) and birthweight was 3.35 kgs (IQR 2.97-3.71). cSO2 increased and FTOE decreased over the first 24 h in all grades of HIE. Compared to the moderate group, infants with mild HIE had significantly higher cSO2 at 6 h (p = 0.003), 9 h (p = 0.009) and 12 h (p = 0.032) and lower FTOE at 6 h (p = 0.016) and 9 h (0.029). cSO2 and FTOE at 6 and 12 h did not predict abnormal outcome. CONCLUSION: Infants with mild HIE have higher cSO2 and lower FTOE than those with moderate or severe HIE in the first 12 h of life. cSO2 increased in all grades of HIE over the first 24 h regardless of TH status.

Electroencephalographic sleep macrostructure and sleep spindles in early infancy.

STUDY OBJECTIVES: Sleep features in infancy are potential biomarkers for brain maturation but poorly characterized. We describe normative values for sleep macrostructure and sleep spindles at 4-5 months of age. METHODS: Healthy term infants were recruited at birth and had daytime sleep electroencephalograms (EEGs) at 4-5 months. Sleep staging was performed and five features were analyzed. Sleep spindles were annotated and seven quantitative features were extracted. Features were analyzed across sex, recording time (am/pm), infant age, and from first to second sleep cycles. RESULTS: We analyzed sleep recordings from 91 infants, 41% females. Median (interquartile range [IQR]) macrostructure results: sleep duration 49.0 (37.8-72.0) min (n = 77); first sleep cycle duration 42.8 (37.0-51.4) min; rapid eye movement (REM) percentage 17.4 (9.5-27.7)% (n = 68); latency to REM 36.0 (30.5-41.1) min (n = 66). First cycle median (IQR) values for spindle features: number 241.0 (193.0-286.5), density 6.6 (5.7-8.0) spindles/min (n = 77); mean frequency 13.0 (12.8-13.3) Hz, mean duration 2.9 (2.6-3.6) s, spectral power 7.8 (4.7-11.4) µV2, brain symmetry index 0.20 (0.16-0.29), synchrony 59.5 (53.2-63.8)% (n = 91). In males, spindle spectral power (µV2) was 24.5% lower (p = .032) and brain symmetry index 24.2% higher than females (p = .011) when controlling for gestational and postnatal age and timing of the nap. We found no other significant associations between studied sleep features and sex, recording time (am/pm), or age. Spectral power decreased (p < .001) on the second cycle. CONCLUSION: This normative data may be useful for comparison with future studies of sleep dysfunction and atypical neurodevelopment in infancy. Clinical Trial Registration: BABY SMART (Study of Massage Therapy, Sleep And neurodevelopMenT) (BabySMART)URL: https://clinicaltrials.gov/ct2/show/results/NCT03381027?view=results.ClinicalTrials.gov Identifier: NCT03381027.

Heart rate variability analysis for the prediction of EEG grade in infants with hypoxic ischaemic encephalopathy within the first 12†h of birth.

Background and aims: Heart rate variability (HRV) has previously been assessed as a biomarker for brain injury and prognosis in neonates. The aim of this cohort study was to use HRV to predict the electroencephalography (EEG) grade in neonatal hypoxic-ischaemic encephalopathy (HIE) within the first 12 h. Methods: We included 120 infants with HIE recruited as part of two European multi-centre studies, with electrocardiography (ECG) and EEG monitoring performed before 12 h of age. HRV features and EEG background were assessed using the earliest 1 h epoch of ECG-EEG monitoring. HRV was expressed in time, frequency and complexity features. EEG background was graded from 0-normal, 1-mild, 2-moderate, 3-major abnormalities to 4-inactive. Clinical parameters known within 6 h of birth were collected (intrapartum complications, foetal distress, gestational age, mode of delivery, gender, birth weight, Apgar at 1 and 5, assisted ventilation at 10 min). Using logistic regression analysis, prediction models for EEG severity were developed for HRV features and clinical parameters, separately and combined. Multivariable model analysis included 101 infants without missing data. Results: Of 120 infants included, 54 (45%) had normal-mild and 66 (55%) had moderate-severe EEG grade. The performance of HRV model was AUROC 0.837 (95% CI: 0.759-0.914) and clinical model was AUROC 0.836 (95% CI: 0.759-0.914). The HRV and clinical model combined had an AUROC of 0.895 (95% CI: 0.832-0.958). Therapeutic hypothermia and anti-seizure medication did not affect the model performance. Conclusions: Early HRV and clinical information accurately predicted EEG grade in HIE within the first 12 h of birth. This might be beneficial when EEG monitoring is not available in the early postnatal period and for referral centres who may want some objective information on HIE severity.

Sparse-Denoising Methods for Extracting Desaturation Transients in Cerebral Oxygenation Signals of Preterm Infants.

Preterm infants are at high risk of developing brain injury in the first days of life as a consequence of poor cerebral oxygen delivery. Near-infrared spectroscopy (NIRS) is an established technology developed to monitor regional tissue oxygenation. Detailed waveform analysis of the cerebral NIRS signal could improve the clinical utility of this method in accurately predicting brain injury. Frequent transient cerebral oxygen desaturations are commonly observed in extremely preterm infants, yet their clinical significance remains unclear. The aim of this study was to examine and compare the performance of two distinct approaches in isolating and extracting transient deflections within NIRS signals. We optimized three different simultaneous low-pass filtering and total variation denoising (LPF-TVD) methods and compared their performance with a recently proposed method that uses singular-spectrum analysis and the discrete cosine transform (SSA-DCT). Parameters for the LPF-TVD methods were optimized over a grid search using synthetic NIRS-like signals. The SSA-DCT method was modified with a post-processing procedure to increase sparsity in the extracted components. Our analysis, using a synthetic NIRS-like dataset, showed that a LPF-TVD method outperformed the modified SSA-DCT method: median mean-squared error of 0.97 (95% CI: 0.86 to 1.07) was lower for the LPF-TVD method compared to the modified SSA-DCT method of 1.48 (95% CI: 1.33 to 1.63), P<0.001. The dual low-pass filter and total variation denoising methods are considerably more computational efficient, by 3 to 4 orders of magnitude, than the SSA-DCT method. More research is needed to examine the efficacy of these methods in extracting oxygen desaturation in real NIRS signals.Clinical relevance- Early and precise identification of abnormal brain oxygenation in premature infants would enable clinicians to employ therapeutic strategies that seek to prevent brain injury and long-term morbidity in this vulnerable population.

In-Silico Evaluation of Anthropomorphic Measurement Variations on Electrical Cardiometry in Neonates.

Non-invasive cardiac output methods such as Electrical Cardiometry (EC) are relatively novel assessment tools for neonates and they enable continuous monitoring of stroke volume (SV). An in-silico comparison of differences in EC-derived SV in relation to preset length and weight was performed. EC (ICON, Osypka Medical) was simulated using the "demo" mode for various combinations of length and weight representative of term and preterm infants. One-centimetre length error resulted in a SV-change of 1.8-3.6% (preterm) or 1.6-2.0% (term) throughout the tested weight ranges. One-hundred gram error in weight measurement resulted in a SV-change of 5.0-7.1% (preterm) or 1.5-1.8% (term) throughout the tested length ranges. Algorithms to calculate EC-derived SV incorporate anthropomorphic measurements. Therefore, inaccuracy in physical measurement can impact absolute EC measurements. This should be considered in the interpretation of previous findings and the design of future clinical studies of EC-derived cardiac parameters in neonates, particularly in the preterm cohorts where a proportional change was noted to be greatest.

Signal Quality of Electrical Cardiometry and Perfusion Index in Very Preterm Infants.

OBJECTIVE: The use of noninvasive monitoring of neonatal hemodynamics is increasing in neonatal care. Methods include noninvasive cardiac output estimated by electrical cardiometry (EC) and peripheral perfusion as perfusion index (PI) using pulse oximetry. Our aim was to evaluate the feasibility to continuously monitor preterm infants with EC and PI during the first 2 postnatal days and the effects of averaging EC data in signal quality (SigQ) analysis. DESIGN: Prospective observational study. SETTING: Tertiary neonatal academic hospital. PATIENTS: Preterm infants <32 weeks gestation from birth until 48 h. MAIN OUTCOME MEASURES: Continuous EC and PI measurements. Feasibility was quantified as the time with high SigQ, classified using SigQ index in EC and exception codes in PI. Our predefined threshold for good feasibility was minimum of 24 h with high SigQ for both. RESULTS: Twenty-two preterm infants (median [IQR] gestational age 28 + 6 (26 + 0, 30 + 4) weeks + days, birth weight 960 [773, 1,500] g) were included. We recorded a minimum of 24 h with high SigQ in 14 infants for EC (unaveraged data) and 22 infants for PI measurements. The median (range) % of recording time with high SigQ was 74% (50%, 88%) for EC and 94% (82%, 96%) for PI. Using 1 minute averaging for EC data resulted in an increase of infants with minimum 24 h of high SigQ to 21 infants. CONCLUSIONS: EC and PI monitoring are feasible in preterm infants within the first 48 h, but SigQ remains problematic for EC. Signal dropout is masked in averaged EC values.

Multichannel EEG abnormalities during the first 6 hours in infants with mild hypoxic-ischaemic encephalopathy.

BACKGROUND: Infants with mild HIE are at risk of significant disability at follow-up. In the pre-therapeutic hypothermia (TH) era, electroencephalography (EEG) within 6 hours of birth was most predictive of outcome. This study aims to identify and describe features of early EEG and heart rate variability (HRV) (<6 hours of age) in infants with mild HIE compared to healthy term infants. METHODS: Infants >36 weeks with mild HIE, not undergoing TH, with EEG before 6 hours of age were identified from 4 prospective cohort studies conducted in the Cork University Maternity Services, Ireland (2003-2019). Control infants were taken from a contemporaneous study examining brain activity in healthy term infants. EEGs were qualitatively analysed by two neonatal neurophysiologists and quantitatively assessed using multiple features of amplitude, spectral shape and inter-hemispheric connectivity. Quantitative features of HRV were assessed in both the groups. RESULTS: Fifty-eight infants with mild HIE and sixteen healthy term infants were included. Seventy-two percent of infants with mild HIE had at least one abnormal EEG feature on qualitative analysis and quantitative EEG analysis revealed significant differences in spectral features between the two groups. HRV analysis did not differentiate between the groups. CONCLUSIONS: Qualitative and quantitative analysis of the EEG before 6 hours of age identified abnormal EEG features in mild HIE, which could aid in the objective identification of cases for future TH trials in mild HIE. IMPACT: Infants with mild HIE currently do not meet selection criteria for TH yet may be at risk of significant disability at follow-up. In the pre-TH era, EEG within 6 hours of birth was most predictive of outcome; however, TH has delayed this predictive value. 72% of infants with mild HIE had at least one abnormal EEG feature in the first 6 hours on qualitative assessment. Quantitative EEG analysis revealed significant differences in spectral features between infants with mild HIE and healthy term infants. Quantitative EEG features may aid in the objective identification of cases for future TH trials in mild HIE.

Cerebral oxygen saturation and autoregulation during hypotension in extremely preterm infants.

BACKGROUND: The impact of the permissive hypotension approach in clinically well infants on regional cerebral oxygen saturation (rScO2) and autoregulatory capacity (CAR) remains unknown. METHODS: Prospective cohort study of blinded rScO2 measurements within a randomized controlled trial of management of hypotension (HIP trial) in extremely preterm infants. rScO2, mean arterial blood pressure, duration of cerebral hypoxia, and transfer function (TF) gain inversely proportional to CAR, were compared between hypotensive infants randomized to receive dopamine or placebo and between hypotensive and non-hypotensive infants, and related to early intraventricular hemorrhage or death. RESULTS: In 89 potentially eligible HIP trial patients with rScO2 measurements, the duration of cerebral hypoxia was significantly higher in 36 hypotensive compared to 53 non-hypotensive infants. In 29/36 hypotensive infants (mean GA 25 weeks, 69% males) receiving the study drug, no significant difference in rScO2 was observed after dopamine (n = 13) compared to placebo (n = 16). Duration of cerebral hypoxia was associated with early intraventricular hemorrhage or death.  Calculated TF gain (n = 49/89) was significantly higher reflecting decreased CAR in 16 hypotensive compared to 33 non-hypotensive infants. CONCLUSIONS: Dopamine had no effect on rScO2 compared to placebo in hypotensive infants. Hypotension and cerebral hypoxia are associated with early intraventricular hemorrhage or death. IMPACT: Treatment of hypotension with dopamine in extremely preterm infants increases mean arterial blood pressure, but does not improve cerebral oxygenation. Hypotensive extremely preterm infants have increased duration of cerebral hypoxia and reduced cerebral autoregulatory capacity compared to non-hypotensive infants. Duration of cerebral hypoxia and hypotension are associated with early intraventricular hemorrhage or death in extremely preterm infants. Since systematic treatment of hypotension may not be associated with better outcomes, the diagnosis of cerebral hypoxia in hypotensive extremely preterm infants might guide treatment.

Can EEG accurately predict 2-year neurodevelopmental outcome for preterm infants?

OBJECTIVE: Establish if serial, multichannel video electroencephalography (EEG) in preterm infants can accurately predict 2-year neurodevelopmental outcome. DESIGN AND PATIENTS: EEGs were recorded at three time points over the neonatal course for infants <32 weeks' gestational age (GA). Monitoring commenced soon after birth and continued over the first 3 days. EEGs were repeated at approximately 32 and 35 weeks' postmenstrual age (PMA). EEG scores were based on an age-specific grading scheme. Clinical score of neonatal morbidity risk and cranial ultrasound imaging were completed. SETTING: Neonatal intensive care unit at Cork University Maternity Hospital, Ireland. MAIN OUTCOME MEASURES: Bayley Scales of Infant Development III at 2 years' corrected age. RESULTS: Sixty-seven infants were prospectively enrolled in the study and 57 had follow-up available (median GA 28.9 weeks (IQR 26.5-30.4)). Forty had normal outcome, 17 had abnormal outcome/died. All EEG time points were individually predictive of abnormal outcome; however, the 35-week EEG performed best. The area under the receiver operating characteristic curve (AUC) for this time point was 0.91 (95% CI 0.83 to 1), p<0.001. Comparatively, the clinical course AUC was 0.68 (95% CI 0.54 to 0.80, p=0.015), while abnormal cranial ultrasound was 0.58 (95% CI 0.41 to 0.75, p=0.342). CONCLUSION: Multichannel EEG is a strong predictor of 2-year outcome in preterm infants particularly when recorded around 35 weeks' PMA. Infants at high risk of brain injury may benefit from early postnatal EEG recording which, if normal, is reassuring. Postnatal clinical complications can contribute to poor outcome; therefore, we state that a later EEG around 35 weeks has a role to play in prognostication.

Grading hypoxic-ischemic encephalopathy in neonatal EEG with convolutional neural networks and quadratic time-frequency distributions.

Objective.To develop an automated system to classify the severity of hypoxic-ischaemic encephalopathy injury (HIE) in neonates from the background electroencephalogram (EEG).Approach. By combining a quadratic time-frequency distribution (TFD) with a convolutional neural network, we develop a system that classifies 4 EEG grades of HIE. The network learns directly from the two-dimensional TFD through 3 independent layers with convolution in the time, frequency, and time-frequency directions. Computationally efficient algorithms make it feasible to transform each 5 min epoch to the time-frequency domain by controlling for oversampling to reduce both computation and computer memory. The system is developed on EEG recordings from 54 neonates. Then the system is validated on a large unseen dataset of 338 h of EEG recordings from 91 neonates obtained across multiple international centres.Main results.The proposed EEG HIE-grading system achieves a leave-one-subject-out testing accuracy of 88.9% and kappa of 0.84 on the development dataset. Accuracy for the large unseen test dataset is 69.5% (95% confidence interval, CI: 65.3%-73.6%) and kappa of 0.54, which is a significant (P<0.001) improvement over a state-of-the-art feature-based method with an accuracy of 56.8% (95% CI: 51.4%-61.7%) and kappa of 0.39. Performance of the proposed system was unaffected when the number of channels in testing was reduced from 8 to 2-accuracy for the large validation dataset remained at 69.5% (95% CI: 65.5%-74.0%).Significance.The proposed system outperforms the state-of-the-art machine learning algorithms for EEG grade classification on a large multi-centre unseen dataset, indicating the potential to assist clinical decision making for neonates with HIE.

Mathematical Analysis of EEG Concordance in Preterm Twin Infants.

PURPOSE: Preterm twins are at higher risk of neurodisability than preterm singletons, with monochorionic-diamniotic (MCDA) twins at higher risk than dichorionic-diamniotic (DCDA) twins. The impact of genetic influences on EEG concordance in preterm twins <32 weeks of gestational age is not established. This study aims to investigate EEG concordance in preterm MCDA and dichorionic-diamniotic twins during maturation. METHODS: Infants <32 weeks of gestational age had multichannel EEG recordings for up to 72 postnatal hours, with repeat recordings at 32 and 35 weeks of postmenstrual age. Twin pairs had synchronous recordings. Mathematical EEG features were generated to represent EEG power, discontinuity, and symmetry. Intraclass correlations, while controlling for gestational age, estimated similarities within twins. RESULTS: EEGs from 10 twin pairs, 4 MCDA and 6 dichorionic-diamniotic pairs, and 10 age-matched singleton pairs were analyzed from a total of 36 preterm infants. For MCDA twins, 17 of 22 mathematical EEG features had significant (>0.6; P < 0.05) intraclass correlations at one or more time points, compared with 2 of 22 features for DCDA twins and 0 of 22 for singleton pairs. For MCDA twins, all 10 features of discontinuity and all four features of symmetry were significant at one or more time-points. Three features of the MCDA twins (spectral power at 3-8 Hz, EEG skewness at 3-15 Hz, and kurtosis at 3-15 Hz) had significant intraclass correlations over all three time points. CONCLUSIONS: Preterm twin EEG similarities are subtle but clearly evident through mathematical analysis. MCDA twins showed stronger EEG concordance across different postmenstrual ages, thus confirming a strong genetic influence on preterm EEG activity at this early development stage.

Detection of Transient Bursts in the EEG of Preterm Infants using Time-Frequency Distributions and Machine Learning.

Short-duration bursts of spontaneous activity are important markers of maturation in the electroencephalogram (EEG) of premature infants. This paper examines the application of a feature-less machine learning approach for detecting these bursts. EEGs were recorded over the first 3 days of life for infants with a gestational age below 30 weeks. Bursts were annotated on the EEG from 36 infants. In place of feature extraction, the time-series EEG is transformed into a time-frequency distribution (TFD). A gradient boosting machine is then trained directly on the whole TFD using a leave-one-out procedure. TFD kernel parameters, length of the Doppler and lag windows, are selected within a nested cross-validation procedure during training. Results indicate that detection performance is sensitive to Doppler-window length but not lag-window length. Median area under the receiver operator characteristic for detection is 0.881 (inter-quartile range 0.850 to 0.913). Examination of feature importance highlights a critical wideband region <15 Hz in the TFD. Burst detection methods form an important component in any fully-automated brain-health index for the vulnerable preterm infant.

Identifying tracé alternant activity in neonatal EEG using an inter-burst detection approach.

Electroencephalography (EEG) is an important clinical tool for reviewing sleep-wake cycling in neonates in intensive care. Tracé alternant (TA)-a characteristic pattern of EEG activity during quiet sleep in term neonates-is defined by alternating periods of short-duration, high-voltage activity (bursts) separated by lower-voltage activity (inter-bursts). This study presents a novel approach for detecting TA activity by first detecting the inter-bursts and then processing the temporal map of the bursts and inter-bursts. EEG recordings from 72 healthy term neonates were used to develop and evaluate performance of 1) an inter-burst detection method which is then used for 2) detection of TA activity. First, multiple amplitude and spectral features were combined using a support vector machine (SVM) to classify bursts from inter-bursts within TA activity, resulting in a median area under the operating characteristic curve (AUC) of 0.95 (95% confidence interval, CI: 0.93 to 0.98). Second, post-processing of the continuous SVM output, the confidence score, was used to produce a TA envelope. This envelope was used to detect TA activity within the continuous EEG with a median AUC of 0.84 (95% CI: 0.80 to 0.88). These results validate how an inter-burst detection approach combined with post processing can be used to classify TA activity. Detecting the presence or absence of TA will help quantify disruption of the clinically important sleep-wake cycle.