Goal: This paper presents a methodology for the automatic detection of sleep apnea from single-lead ECG. Methods: It uses two novel features derived from the ECG, and two well-known features in heart rate variability analysis, namely the standard deviation and the serial correlation coefficients of the RR interval time series. The first novel feature uses the principal components of the QRS complexes, and it describes changes in their morphology caused by an increased sympathetic activity during apnea. The second novel feature extracts the information shared between respiration and heart rate using orthogonal subspace projections. Respiratory information is derived from the ECG by means of three state-of-the-art algorithms, which are implemented and compared here. All features are used as input to a least-squares support vector machines classifier, using an RBF kernel. In total, 80 ECG recordings were included in the study. Results: Accuracies of about 85% are achieved on a minute-by-minute basis, for two independent datasets including both hypopneas and apneas together. Separation between apnea and normal recordings is achieved with 100% accuracy. In addition to apnea classification, the proposed methodology determines the contamination level of each ECG minute. Conclusion: The performances achieved are comparable with those reported in the literature for fully automated algorithms. Significance: These results indicate that the use of only ECG sensors can achieve good accuracies in the detection of sleep apnea. Moreover, the contamination level of each ECG segment can be used to automatically detect artefacts, and to highlight segments that require further visual inspection.

Identifying a core set of features is one of the most important steps in the development of an automated seizure detector. In most of the published studies describing features and seizure classifiers, the features were hand-engineered, which may not be optimal. The main goal of the present paper is using deep convolutional neural networks (CNNs) and random forest to automatically optimize feature selection and classification. The input of the proposed classifier is raw multi-channel EEG and the output is the class label: seizure/nonseizure. By training this network, the required features are optimized, while fitting a nonlinear classifier on the features. After training the network with EEG recordings of 26 neonates, five end layers performing the classification were replaced with a random forest classifier in order to improve the performance. This resulted in a false alarm rate of 0.9 per hour and seizure detection rate of 77% using a test set of EEG recordings of 22 neonates that also included dubious seizures. The newly proposed CNN classifier outperformed three data-driven feature-based approaches and performed similar to a previously developed heuristic method.

Currently, reliable reference values of regional cerebral oxygen saturation (rScO₂) for different gestational age (GA) groups are lacking, which hampers the implementation of near-infrared spectroscopy (NIRS) alongside monitoring arterial oxygen saturation (SaO₂) and blood pressure in neonatal intensive care. The aim of this study was to provide reference values for rScO₂ and cerebral fractional tissue oxygen extraction (cFTOE; (SaO₂ − rScO₂)/SaO₂) for small adult and neonatal NIRS sensors. In this study, 999 infants born preterm (GA <32 wk) were monitored with NIRS during the first 72 h of life. Mixed modeling was used to generate reference curves grouped per 2 wk of GA. In addition, the influence of a hemodynamically significant patent ductus arteriosus, gender, and birth weight were explored. Average rScO₂ was ~65% at admission, increased with GA (1% per week) and followed a parabolic curve in relation to postnatal age with a peak at ~36 h. The cFTOE showed similar but inverse effects. On average, the neonatal sensor measured 10% higher than the adult sensor. rScO₂ and cFTOE reference curves are provided for the first 72 h of life in preterm infants, which might support the broader implementation of NIRS in neonatal intensive care.

Introduction: Cerebral autoregulation (CAR), the ability of the human body to maintain cerebral blood flow (CBF) in a wide range of perfusion pressures, can be calculated by describing the relation between arterial blood pressure (ABP) and cerebral oxygen saturation measured by near-infrared spectroscopy (NIRS). In literature, disturbed CAR is described in different patient groups, using multiple measurement techniques and mathematical models. Furthermore, it is unclear to what extent cerebral pathology and outcome can be explained by impaired CAR. Aim and methods: In order to summarize CAR studies using NIRS in neonates, a systematic review was performed in the PUBMED and EMBASE database. To provide a general overview of the clinical framework used to study CAR, the different preprocessing methods and mathematical models are described and explained. Furthermore, patient characteristics, definition of impaired CAR and the outcome according to this definition is described organized for the different patient groups. Results: Forty-six articles were included in this review. Four patient groups were established: preterm infants during the transitional period, neonates receiving specific medication/treatment, neonates with congenital heart disease and neonates with hypoxic-ischemic encephalopathy (HIE) treated with therapeutic hypothermia. Correlation, coherence and transfer function (TF) gain are the mathematical models most frequently used to describe CAR. The definition of impaired CAR is depending on the mathematical model used. The incidence of intraventricular hemorrhage in preterm infants is the outcome variable most frequently correlated with impaired CAR. Hypotension, disease severity, dopamine treatment, injury on magnetic resonance imaging (MRI) and long term outcome are associated with impaired CAR. Prospective interventional studies are lacking in all research areas. Discussion and conclusion: NIRS derived CAR measurement is an important research tool to improve knowledge about central hemodynamic fluctuations during the transitional period, cerebral pharmacodynamics of frequently used medication (sedatives-inotropes) and cerebral effects of specific therapies in neonatology. Uniformity regarding measurement techniques and mathematical models is needed. Multimodal monitoring databases of neonatal intensive care patients of multiple centers, together with identical outcome parameters are needed to compare different techniques and make progress in this field. Real-time bedside monitoring of CAR, together with conventional monitoring, seems a promising technique to improve individual patient care.