4.6 Article

Improved multi-stage neonatal seizure detection using a heuristic classifier and a data-driven post-processor

Journal

CLINICAL NEUROPHYSIOLOGY
Volume 127, Issue 9, Pages 3014-3024

Publisher

ELSEVIER IRELAND LTD
DOI: 10.1016/j.clinph.2016.06.018

Keywords

Automated neonatal seizure detection; Hypoxic-ischemic encephalopathy; Machine learning; Support vector machines

Funding

  1. Bijzonder Onderzoeksfonds KU Leuven (BOF): Center of Excellence (CoE) [PFV/10/002]
  2. Fonds voor Wetenschappelijk Onderzoek-Vlaanderen (FWO) [G.0427.10N, G.0108.11, G.0869.12N, G.0A5513N]
  3. Agentschap voor Innovatie door Wetenschap en Technologie (IWT) [TBM 080658-MRI, TBM 110697-NeoGuard]
  4. iMinds Medical Information Technologies: Dotatie-Strategisch basis onderzoek (SBO), ICON: NXT_Sleep
  5. Belgian Federal Science Policy Office [IUAP P7/19/]
  6. Belgian Foreign Affairs-Development Cooperation: VLIR UOS programs
  7. European Union Seventh Framework Programme (FP7): EU MC ITN TRANSACT [316679]
  8. European Union Seventh Framework Programme (FP7): ERASMUS EQR: Community service engineer [539642-LLP-1-2013]
  9. EU INTERREG IVB NWE programme [RECAP 209G]
  10. European Research Council (ERC) [339804]
  11. IWT PHD grant [TBM 110697-NeoGuard]

Ask authors/readers for more resources

Objective: After identifying the most seizure-relevant characteristics by a previously developed heuristic classifier, a data-driven post-processor using a novel set of features is applied to improve the performance. Methods: The main characteristics of the outputs of the heuristic algorithm are extracted by five sets of features including synchronization, evolution, retention, segment, and signal features. Then, a support vector machine and a decision making layer remove the falsely detected segments. Results: Four datasets including 71 neonates (1023 h, 3493 seizures) recorded in two different university hospitals, are used to train and test the algorithm without removing the dubious seizures. The heuristic method resulted in a false alarm rate of 3.81 per hour and good detection rate of 88% on the entire test databases. The post-processor, effectively reduces the false alarm rate by 34% while the good detection rate decreases by 2%. Conclusion: This post-processing technique improves the performance of the heuristic algorithm. The structure of this post-processor is generic, improves our understanding of the core visually determined EEG features of neonatal seizures and is applicable for other neonatal seizure detectors. Significance: The post-processor significantly decreases the false alarm rate at the expense of a small reduction of the good detection rate. (C) 2016 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.6
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

Article Medicine, General & Internal

Preterm birth during the COVID-19 pandemic: more, less, or just the same?

Isabelle Dehaene, Caroline Van Holsbeke, Kristien Roelens, Noortje Nvh van Oostrum, Katrien Nulens, Koenraad Smets, Marie-Rose Van Hoestenberghe, Gunnar Naulaers, Roland Devlieger

Summary: The impact of COVID-19 and its mitigation measures on preterm birth incidence is still unclear and varies across countries. In three Belgian tertiary care hospitals, there were no differences in preterm birth rates between 2019 and 2020.

ACTA CLINICA BELGICA (2023)

Review Pediatrics

Cerebral oxygenation and body position in the preterm infant: A systematic review and meta-analysis

Eva Maes, Filip Cools, Anneleen Dereymaeker, Katrien Jansen, Gunnar Naulaers, Liesbeth Thewissen

Summary: This systematic review and meta-analysis investigated the effect of body position on rScO2 in preterm infants during the first 2 weeks after birth. The results showed that the prone position led to significantly higher rScO2 in the second week after birth compared to the supine position, but no difference was observed in extremely preterm and preterm infants.

ACTA PAEDIATRICA (2023)

Letter Respiratory System

Implications of the new ERS/ATS standards on the interpretation of lung function tests

Paul Desbordes, Maarten De Vos, Julie Maes, Frans de Jongh, Karl Sylvester, Claus Franz Vogelmeier, Anh Tuan Dinh-Xuan, Jann Mortensen, Wim Janssens, Marko Topalovic

EUROPEAN RESPIRATORY JOURNAL (2023)

Article Engineering, Biomedical

A multi-task and multi-channel convolutional neural network for semi-supervised neonatal artefact detection

Tim Hermans, Laura Smets, Katrien Lemmens, Anneleen Dereymaeker, Katrien Jansen, Gunnar Naulaers, Filippo Zappasodi, Sabine Van Huffel, Silvia Comani, Maarten De Vos

Summary: This paper proposes a semi-supervised deep learning approach for artefact detection in neonatal EEG. The proposed method outperforms existing state-of-the-art models and achieves good performance on two separate datasets. The results demonstrate the effectiveness of the semi-supervised multi-task training strategy and the relevance of artefact detection for automated EEG analysis.

JOURNAL OF NEURAL ENGINEERING (2023)

Article Clinical Neurology

Investigating the development of the autonomic nervous system in infancy through pupillometry

Lyssa de Vries, Steffie Amelynck, Paer Nystroem, Lotte van Esch, Thijs Van Lierde, Petra Warreyn, Herbert Roeyers, Ilse Noens, Gunnar Naulaers, Bart Boets, Jean Steyaert, TIARA Team

Summary: This study aimed to investigate the developmental trajectories of the autonomic nervous system (ANS) in infants, with or without increased likelihood for atypical ANS development. The pupillary light reflex (PLR) was used as an index and eye-tracking was employed to capture the PLR in 216 infants. The results revealed an increase in PLR parameters with age, as well as group differences in baseline pupil diameter and latency to constriction.

JOURNAL OF NEURAL TRANSMISSION (2023)

Article Biotechnology & Applied Microbiology

The Power of ECG in Semi-Automated Seizure Detection in Addition to Two-Channel behind-the-Ear EEG

Miguel Bhagubai, Kaat Vandecasteele, Lauren Swinnen, Jaiver Macea, Christos Chatzichristos, Maarten De Vos, Wim Van Paesschen

Summary: This study evaluated a semi-automated multimodal wearable seizure detection framework using bte-EEG and ECG data. The results showed that combining ECG with bte-EEG can improve the accuracy of seizure detection and reduce false alarm rates, while also saving time for both clinicians and patients.

BIOENGINEERING-BASEL (2023)

Article Clinical Neurology

Microstate Analysis Reflects Maturation of the Preterm Brain

Tim Hermans, Mohammad Khazaei, Khadijeh Raeisi, Pierpaolo Croce, Gabriella Tamburro, Anneleen Dereymaeker, Maarten De Vos, Filippo Zappasodi, Silvia Comani

Summary: This study aims to explore the use of microstate (MS) analysis to evaluate global brain dynamics changes during maturation in preterm neonates with normal neurodevelopmental outcome. The study found that MS duration decreased and occurrence increased with age in preterm neonates, and significant changes in MS topographies and transitions occurred when neonates reached 37 weeks. Additionally, the Hurst exponent of the individual MS sequence decreased with age.

BRAIN TOPOGRAPHY (2023)

Article Clinical Neurology

Longitudinal trajectory of medial gastrocnemius muscle growth in the first years of life

Nathalie De Beukelaer, Ines Vandekerckhove, Geert Molenberghs, Gunnar Naulaers, Liesbeth Thewissen, Domiziana Costamagna, Anja Van Campenhout, Kaat Desloovere, Els Ortibus

Summary: This study found that muscle growth is already affected in infants with spastic cerebral palsy. Muscle size growth further reduces with decreasing functional levels, regardless of the brain lesion. Early monitoring and intervention for muscle growth is necessary.

DEVELOPMENTAL MEDICINE AND CHILD NEUROLOGY (2023)

Article Pediatrics

The Impact of Intraventricular Hemorrhage and Periventricular Leukomalacia on Mortality and Neurodevelopmental Outcome in Very Preterm and Very Low Birthweight Infants: A Prospective Population-based Cohort Study

Aurelie Pascal, Nele de Bruyn, Gunnar Naulaers, Els Ortibus, Britta Hanssen, Ann Oostra, Kris de Coen, Michel Sonnaert, Eva Cloet, Alexandra Casaer, James D'Haese, Sabine Laroche, An Jonckheere, Katleen Plaskie, Christine van Mol, Els Bruneel, Marie-Rose van Hoestenberghe, Bieke Samijn, Paul Govaert, Christine Van den Broeck

Summary: This study investigated the incidence of intraventricular hemorrhage (IVH) and periventricular leukomalacia (PVL) in relation to gestational age, as well as the impact on mortality and neurodevelopmental outcomes in very preterm/very low birthweight infants. The results showed that both the prevalence and severity of IVH/PVL decreased significantly as gestational age advanced. More than 75% of infants with low grades of IVH/PVL had normal motor and cognitive outcomes at 2 years of age. High-grade PVL/IVH has become less common and is associated with adverse outcomes.

JOURNAL OF PEDIATRICS (2023)

Meeting Abstract Neurosciences

The BELpREG registration system on perinatal medication use and mother-infant outcomes: preliminary insights into the cohort's characteristics (Presented at the ENTIS business meeting)

Laure Sillisa, Cleo Vandegoor, Cato Vercaeren, Karel Allegaert, Annick Bogaerts, Maarten De Vos, Titia Hompes, Anne Smits, Kristel Van Calsteren, Jan Y. Verbakel, Veerle Foulon, Michael Ceulemans

NEUROTOXICOLOGY AND TERATOLOGY (2023)

Meeting Abstract Critical Care Medicine

How Many Lung Function Interpretations Can an Expert Make in One Hour Based on the 2021 ERS/ATS Guidelines?

M. Topalovic, M. De Vos, J. Maes, J. Kaspers, N. Stachowicz, P. Desbordes

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE (2023)

Meeting Abstract Transplantation

AUTOMATIC SEGMENTATION OF GLOMERULAR SUBSTRUCTURES BY DEEP LEARNING

Amelie Dendooven, Aristotelis Styanidis, Louis Raes, Amaryllis Van Craenenbroeck, Matthias Maeyens, Konstantinos Kotras, Maarten De Vos

NEPHROLOGY DIALYSIS TRANSPLANTATION (2023)

Article Respiratory System

Principal component analysis of flow-volume curves in COPDGene to link spirometry with phenotypes of COPD

Kenneth Verstraete, Nilakash Das, Iwein Gyselinck, Marko Topalovic, Thierry Troosters, James D. Crapo, Edwin K. Silverman, Barry J. Make, Elizabeth A. Regan, Robert Jensen, Maarten De Vos, Wim Janssens

Summary: The shape of MEFVC is associated with CT parameters of emphysema, small airways disease (SAD), and bronchial wall thickening (BWT) in COPD. It is a valuable predictor for emphysema and SAD in moderate-severe COPD, but not a suitable screening tool for early disease phenotypes identified by CT scan.

RESPIRATORY RESEARCH (2023)

Proceedings Paper Computer Science, Artificial Intelligence

The Statistical Analysis of the Varying Brain

Oliver Y. Chen, Vu Duy Thanh, Gilbert Greub, Hengyi Cao, Xingru He, Yannick Muller, Constantinos Petrovas, Haochang Shou, Viet-Dung Nguyen, Bangdong Zhi, Laurent Perez, Jean-Louis Raisaro, Guy Nagels, Maarten de Vos, Wei He, Gottardo Palie Smart, Marcus Munafo, Giuseppe Pantaleo

Summary: This article presents a systematic approach to studying varying brain. It discusses different types of brain variability and provides examples for each. It explores classical analysis of covariance as well as advanced residual analysis methods that aim to decompose the total variance of brain or behavior data. The article also considers innate and acquired brain variability, the neural law of large numbers for big brain data, and the gut-brain axis as an important source of brain variability.

2023 IEEE STATISTICAL SIGNAL PROCESSING WORKSHOP, SSP (2023)

Article Clinical Neurology

Functional connectivity of sensorimotor network is enhanced in spastic diplegic cerebral palsy: A multimodal study using fMRI and MEG

Jaakko Vallinoja, Timo Nurmi, Julia Jaatela, Vincent Wens, Mathieu Bourguignon, Helena Maenpaa, Harri Piitulainen

Summary: The study aimed to assess the effects of lesions related to spastic diplegic cerebral palsy on functional connectivity. Using multiple imaging modalities, the researchers found enhanced functional connectivity in the sensorimotor network of individuals with spastic diplegic cerebral palsy, which was not correlated with hand coordination performance.

CLINICAL NEUROPHYSIOLOGY (2024)

Article Clinical Neurology

Sensorimotor integration in cranial muscles tested by short- and long-latency afferent inhibition

Francesca Ginatempo, Nicola Loi, John C. Rothwell, Franca Deriu

Summary: This study comprehensively investigated sensorimotor integration in the cranial-cervical muscles of healthy adults and found that the integration of sensory inputs with motor output is profoundly influenced by the type of sensory afferent involved and the functional role played by the target muscle.

CLINICAL NEUROPHYSIOLOGY (2024)