TY - JOUR
T1 - Quantitative Assessment of Electroencephalogram Reactivity in Comatose Patients on Extracorporeal Membrane Oxygenation
AU - Williams, Autumn
AU - Zeng, Yinuo
AU - Li, Ziwei
AU - Thakor, Nitish
AU - Geocadin, Romergryko G.
AU - Bronder, Jay
AU - Martinez, Nirma Carballido
AU - Ritzl, Eva K.
AU - Cho, Sung Min
N1 - Publisher Copyright:
© 2022 World Scientific Publishing Company.
PY - 2022/6/1
Y1 - 2022/6/1
N2 - Objective assessment of the brain's responsiveness in comatose patients on Extracorporeal Membrane Oxygenation (ECMO) support is essential to clinical care, but current approaches are limited by subjective methodology and inter-rater disagreement. Quantitative electroencephalogram (EEG) algorithms could potentially assist clinicians, improving diagnostic accuracy. We developed a quantitative, stimulus-based algorithm to assess EEG reactivity features in comatose patients on ECMO support. Patients underwent a stimulation protocol of increasing intensity (auditory, peripheral, and nostril stimulation). A total of 129 20-s EEG epochs were collected from 24 patients (age 56.9 ± 15.1, 10 females, 14 males) on ECMO support with a Glasgow Coma Scale<8. EEG reactivity scores (R-scores) were calculated using aggregated spectral power and permutation entropy for each of five frequency bands (δ, α, β, γ). Parameter estimation techniques were applied to R-scores to identify properties that replicate the decision process of experienced clinicians performing visual analysis. Spectral power changes from audio stimulation were concentrated in the β band, whereas peripheral stimulation elicited an increase in spectral power across multiple bands, and nostril stimulation changed the entropy of the γ band. The findings of this pilot study on R-score lay a foundation for a future prediction tool with clinical applications.
AB - Objective assessment of the brain's responsiveness in comatose patients on Extracorporeal Membrane Oxygenation (ECMO) support is essential to clinical care, but current approaches are limited by subjective methodology and inter-rater disagreement. Quantitative electroencephalogram (EEG) algorithms could potentially assist clinicians, improving diagnostic accuracy. We developed a quantitative, stimulus-based algorithm to assess EEG reactivity features in comatose patients on ECMO support. Patients underwent a stimulation protocol of increasing intensity (auditory, peripheral, and nostril stimulation). A total of 129 20-s EEG epochs were collected from 24 patients (age 56.9 ± 15.1, 10 females, 14 males) on ECMO support with a Glasgow Coma Scale<8. EEG reactivity scores (R-scores) were calculated using aggregated spectral power and permutation entropy for each of five frequency bands (δ, α, β, γ). Parameter estimation techniques were applied to R-scores to identify properties that replicate the decision process of experienced clinicians performing visual analysis. Spectral power changes from audio stimulation were concentrated in the β band, whereas peripheral stimulation elicited an increase in spectral power across multiple bands, and nostril stimulation changed the entropy of the γ band. The findings of this pilot study on R-score lay a foundation for a future prediction tool with clinical applications.
KW - Coma
KW - Disorder of consciousness
KW - ECMO
KW - EEG reactivity
KW - Quantitative EEG
KW - Regression analysis
UR - http://www.scopus.com/inward/record.url?scp=85129024253&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85129024253&partnerID=8YFLogxK
U2 - 10.1142/S0129065722500253
DO - 10.1142/S0129065722500253
M3 - Article
C2 - 35443895
AN - SCOPUS:85129024253
SN - 0129-0657
VL - 32
JO - International journal of neural systems
JF - International journal of neural systems
IS - 6
M1 - 2250025
ER -