Show simple item record

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

dc.contributor.authorRaurale, Sumit A
dc.contributor.authorBoylan, Geraldine B
dc.contributor.authorMathieson, Sean R
dc.contributor.authorMarnane, William P
dc.contributor.authorLightbody, Gordon
dc.contributor.authorO'Toole, John M
dc.date.accessioned2025-01-31T12:13:29Z
dc.date.available2025-01-31T12:13:29Z
dc.date.issued2021-03-19
dc.identifier.pmid33618337
dc.identifier.doi10.1088/1741-2552/abe8ae
dc.identifier.urihttp://hdl.handle.net/10147/644129
dc.descriptionObjective.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.en_US
dc.language.isoenen_US
dc.rightsCreative Commons Attribution license.
dc.rightsAttribution 4.0 International*
dc.rights.urihttp://creativecommons.org/licenses/by/4.0/*
dc.subjectconvolutional neural networken_US
dc.subjectelectroencephalographyen_US
dc.subjecthypoxic-ischemic encephalopathyen_US
dc.subjecttime-frequency distributionen_US
dc.titleGrading hypoxic-ischemic encephalopathy in neonatal EEG with convolutional neural networks and quadratic time-frequency distributions.en_US
dc.typeArticleen_US
dc.typeOtheren_US
dc.identifier.eissn1741-2552
dc.identifier.journalJournal of neural engineeringen_US
dc.source.journaltitleJournal of neural engineering
dc.source.volume18
dc.source.issue4
refterms.dateFOA2025-01-31T12:13:30Z
dc.source.countryUnited Kingdom
dc.source.countryUnited Kingdom
dc.source.countryEngland


Files in this item

Thumbnail
Name:
jne_18_4_046007.pdf
Size:
3.637Mb
Format:
PDF
Download

This item appears in the following Collection(s)

Show simple item record

Creative Commons Attribution license.
Except where otherwise noted, this item's license is described as Creative Commons Attribution license.