Alterations in coordinated EEG activity precede the development of seizures in comatose children

Vakorin, Vasily A.; Nita, Dragos A.; Payne, Eric T.; McBain, Kristin; Frndova, Helena; Go, Cristina; Ribary, Urs; Abend, Nicholas S.; Gallentine, William B.; Nash, Kendall; Hutchison, James S.; Parshuram, Christopher S.; Snead, O. Carter; Straaten, Ilse E C W van; Stam, Cornelis J.; Doesburg, Sam M.; Hahn, Cecil D.

doi:10.1016/j.clinph.2021.03.015

“…The resulting PPV of the Age+HRV+MOR model was 41% and the PPV of the META+HRV+MOR model was 51%. In other words, the Age+HRV+MOR and the META+HRV+MOR models achieved a relative improvement of 28% and 59% respectively compared to the best clinical practice PPV as reported in [9]. Fig.…”

Section: Usage Of the Developed Algorithms And Clinical Scenariomentioning

confidence: 61%

“…The database includes synchronous cEEG and ECG recordings of 176 children with a reduced level of consciousness. More details regarding this database were presented in a previous study [9] and data usage was granted under SickKids Research Ethics Board number 1000014179. Among these patients, 52 children experienced seizures of any type and any length and 124 children were without any seizures.…”

Section: Database Descriptionmentioning

confidence: 99%

“…In this setting, a random distribution of the 8 cEEG systems over the patients leads to a PPV of 11%. In the current medical examination pathway, among the children suspected of epileptic seizures and monitored using cEEG, only 9-32% of them are reported as presenting epileptic seizure events [9]. Thus, the best reported PPV of the clinical risk stratification system is 32%.…”

Section: Usage Of the Developed Algorithms And Clinical Scenariomentioning

confidence: 99%

“…The use of cEEG has grown in the PICU and became a standard of care in many North American healthcare centers [7]. Among the children suspected of epileptic seizures and monitored using cEEG, status epilepticus, defined as a seizure that lasts more than 5 minutes [8], is established for 9-32% [9] of the monitored patients. In addition, other patients that are not suspected and not monitored with cEEG might be missed.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Machine Learning to Support Triage of Children at Risk for Epileptic Seizures in the Pediatric Intensive Care Unit

Azriel,

Hahn,

De Cooman

et al. 2022

Preprint

Self Cite

0

View full text Add to dashboard Cite

Objective: Epileptic seizures are relatively common in critically-ill children admitted to the pediatric intensive care unit (PICU) and thus serve as an important target for identification and treatment. Most of these seizures have no discernible clinical manifestation but still have a significant impact on morbidity and mortality. Children that are deemed at risk for seizures within the PICU are monitored using continuous-electroencephalogram (cEEG). cEEG monitoring cost is considerable and as the number of available machines is always limited, clinicians need to resort to triaging patients according to perceived risk in order to allocate resources. This research aims to develop a computer aided tool to improve seizures risk assessment in critically-ill children, using an ubiquitously recorded signal in the PICU, namely the electrocardiogram (ECG). Approach: A novel data-driven model was developed at a patient-level approach, based on features extracted from the first hour of ECG recording and the clinical data of the patient. Main results: The most predictive features were the age of the patient, the brain injury as coma etiology and the QRS area. For patients without any prior clinical data, using one hour of ECG recording, the classification performance of the random forest classifier reached an area under the receiver operating characteristic curve (AUROC) score of 0.84. When combining ECG features with the patients clinical history, the AUROC reached 0.87. Significance: Taking a real clinical scenario, we estimated that our clinical decision support triage tool can improve the positive predictive value by more than 59% over the clinical standard.

show abstract

“…This is a highly heterogeneous population, with various diagnoses and a full spectrum of comorbidity levels. These patients were expected to take various medications, which is known to impact EEG dynamics [36]. Furthermore, we did not separate EEG scans into normal and abnormal categories.…”

Section: Discussionmentioning

confidence: 99%

Byte-Pair Encoding for classifying routine clinical electroencephalograms in adults over the lifespan

Klymenko

¹

,

Doesburg

²

,

Medvedev

³

et al. 2023

IEEE J. Biomed. Health Inform.

2

0

View full text Add to dashboard Cite

Machine learning to support triage of children at risk for epileptic seizures in the pediatric intensive care unit

Azriel

¹

,

Hahn

²

,

Cooman

³

et al. 2022

Physiol. Meas.

1

0

View full text Add to dashboard Cite

Objective: Epileptic seizures are relatively common in critically-ill children admitted to the pediatric intensive care unit (PICU) and thus serve as an important target for identification and treatment. Most of these seizures have no discernible clinical manifestation but still have a significant impact on morbidity and mortality. Children that are deemed at risk for seizures within the PICU are monitored using continuous-electroencephalogram (cEEG). cEEG monitoring cost is considerable and as the number of available machines is always limited, clinicians need to resort to triaging patients according to perceived risk in order to allocate resources. This research aims to develop a computer aided tool to improve seizures risk assessment in critically-ill children, using an ubiquitously recorded signal in the PICU, namely the electrocardiogram (ECG). Approach: A novel data-driven model was developed at a patient-level approach, based on features extracted from the first hour of ECG recording and the clinical data of the patient. Main results: The most predictive features were the age of the patient, the brain injury as coma etiology and the QRS area. For patients without any prior clinical data, using one hour of ECG recording, the classification performance of the random forest classifier reached an area under the receiver operating characteristic curve (AUROC) score of 0.84. When combining ECG features with the patients clinical history, the AUROC reached 0.87. Significance: Taking a real clinical scenario, we estimated that our clinical decision support triage tool can improve the positive predictive value by more than 59% over the clinical standard.

show abstract

Alterations in coordinated EEG activity precede the development of seizures in comatose children

Cited by 4 publications

References 55 publications

Machine Learning to Support Triage of Children at Risk for Epileptic Seizures in the Pediatric Intensive Care Unit

Machine Learning to Support Triage of Children at Risk for Epileptic Seizures in the Pediatric Intensive Care Unit

Byte-Pair Encoding for classifying routine clinical electroencephalograms in adults over the lifespan

Machine learning to support triage of children at risk for epileptic seizures in the pediatric intensive care unit

Contact Info

Product

Resources

About