The prognostic value of resting-state EEG in acute post-traumatic unresponsive states

Research output: Contribution to journalArticlepeer-review

Standard

The prognostic value of resting-state EEG in acute post-traumatic unresponsive states. / O'Donnell, Alice; Pauli, Ruth; Banellis, Leah; Sokoliuk, Rodika; Hayton, Tom; Sturman, Steve; Veenith, T; Yakoub, Kamal; Belli, Tony; Chennu, Srivas; Cruse, Damian.

In: Brain Communications, Vol. 3, No. 2, fcab017, 17.03.2021.

Research output: Contribution to journalArticlepeer-review

Harvard

APA

Vancouver

Author

Bibtex

@article{c8cda88761bc40fcb5d40d7ec25deb03,
title = "The prognostic value of resting-state EEG in acute post-traumatic unresponsive states",
abstract = "Accurate early prognostication is vital for appropriate long-term care decisions after traumatic brain injury. While measures of resting-state EEG oscillations and their network properties, derived from graph theory, have been shown to provide clinically useful information regarding diagnosis and recovery in patients with chronic disorders of consciousness, little is known about the value of these network measures when calculated from a standard clinical low-density EEG in the acute phase post-injury. To investigate this link, we first validated a set of measures of oscillatory network features between high-density and low-density resting-state EEG in healthy individuals, thus ensuring accurate estimation of underlying cortical function in clinical recordings from patients. Next, we investigated the relationship between these features and the clinical picture and outcome of a group of 18 patients in acute post-traumatic unresponsive states who were not following commands 2 days+ after sedation hold. While the complexity of the alpha network, as indexed by the standard deviation of the participation coefficients, was significantly related to the patients{\textquoteright} clinical picture at the time of EEG, no network features were significantly related to outcome at 3 or 6 months post-injury. Rather, mean relative alpha power across all electrodes improved the accuracy of outcome prediction at 3 months relative to clinical features alone. These results highlight the link between the alpha rhythm and clinical signs of consciousness and suggest the potential for simple measures of resting-state EEG band power to provide a coarse snapshot of brain health for stratification of patients for rehabilitation, therapy and assessments of both covert and overt cognition.",
author = "Alice O'Donnell and Ruth Pauli and Leah Banellis and Rodika Sokoliuk and Tom Hayton and Steve Sturman and T Veenith and Kamal Yakoub and Tony Belli and Srivas Chennu and Damian Cruse",
year = "2021",
month = mar,
day = "17",
doi = "10.1093/braincomms/fcab017",
language = "English",
volume = "3",
journal = "Brain Communications",
issn = "2632-1297",
publisher = "Oxford University Press",
number = "2",

}

RIS

TY - JOUR

T1 - The prognostic value of resting-state EEG in acute post-traumatic unresponsive states

AU - O'Donnell, Alice

AU - Pauli, Ruth

AU - Banellis, Leah

AU - Sokoliuk, Rodika

AU - Hayton, Tom

AU - Sturman, Steve

AU - Veenith, T

AU - Yakoub, Kamal

AU - Belli, Tony

AU - Chennu, Srivas

AU - Cruse, Damian

PY - 2021/3/17

Y1 - 2021/3/17

N2 - Accurate early prognostication is vital for appropriate long-term care decisions after traumatic brain injury. While measures of resting-state EEG oscillations and their network properties, derived from graph theory, have been shown to provide clinically useful information regarding diagnosis and recovery in patients with chronic disorders of consciousness, little is known about the value of these network measures when calculated from a standard clinical low-density EEG in the acute phase post-injury. To investigate this link, we first validated a set of measures of oscillatory network features between high-density and low-density resting-state EEG in healthy individuals, thus ensuring accurate estimation of underlying cortical function in clinical recordings from patients. Next, we investigated the relationship between these features and the clinical picture and outcome of a group of 18 patients in acute post-traumatic unresponsive states who were not following commands 2 days+ after sedation hold. While the complexity of the alpha network, as indexed by the standard deviation of the participation coefficients, was significantly related to the patients’ clinical picture at the time of EEG, no network features were significantly related to outcome at 3 or 6 months post-injury. Rather, mean relative alpha power across all electrodes improved the accuracy of outcome prediction at 3 months relative to clinical features alone. These results highlight the link between the alpha rhythm and clinical signs of consciousness and suggest the potential for simple measures of resting-state EEG band power to provide a coarse snapshot of brain health for stratification of patients for rehabilitation, therapy and assessments of both covert and overt cognition.

AB - Accurate early prognostication is vital for appropriate long-term care decisions after traumatic brain injury. While measures of resting-state EEG oscillations and their network properties, derived from graph theory, have been shown to provide clinically useful information regarding diagnosis and recovery in patients with chronic disorders of consciousness, little is known about the value of these network measures when calculated from a standard clinical low-density EEG in the acute phase post-injury. To investigate this link, we first validated a set of measures of oscillatory network features between high-density and low-density resting-state EEG in healthy individuals, thus ensuring accurate estimation of underlying cortical function in clinical recordings from patients. Next, we investigated the relationship between these features and the clinical picture and outcome of a group of 18 patients in acute post-traumatic unresponsive states who were not following commands 2 days+ after sedation hold. While the complexity of the alpha network, as indexed by the standard deviation of the participation coefficients, was significantly related to the patients’ clinical picture at the time of EEG, no network features were significantly related to outcome at 3 or 6 months post-injury. Rather, mean relative alpha power across all electrodes improved the accuracy of outcome prediction at 3 months relative to clinical features alone. These results highlight the link between the alpha rhythm and clinical signs of consciousness and suggest the potential for simple measures of resting-state EEG band power to provide a coarse snapshot of brain health for stratification of patients for rehabilitation, therapy and assessments of both covert and overt cognition.

U2 - 10.1093/braincomms/fcab017

DO - 10.1093/braincomms/fcab017

M3 - Article

C2 - 33855295

VL - 3

JO - Brain Communications

JF - Brain Communications

SN - 2632-1297

IS - 2

M1 - fcab017

ER -