Predicting posttraumatic epilepsy using admission electroencephalography after severe traumatic brain injury

Area of Science:

• Neuroscience
• Neurology
• Medical Technology

Background:

• Posttraumatic epilepsy (PTE) affects up to one-third of severe traumatic brain injury (TBI) patients, often manifesting years after the initial injury.
• Early identification of high-risk individuals is crucial for timely clinical management and potential enrollment in clinical trials.

Purpose of the Study:

• To investigate the utility of early electroencephalographic (EEG) features, analyzed via quantitative EEG (qEEG), in predicting the long-term risk of developing PTE after severe TBI.
• To develop predictive models for PTE risk using both qEEG and clinical data.

Main Methods:

• A case-control study involving severe TBI patients, comparing those with and without PTE.
• Continuous EEG monitoring for 3-5 days post-injury, with analysis of a 5-minute early epoch using qEEG.
• Development of multivariable models (random forest and logistic regression) to predict PTE risk.

Main Results:

• Quantitative EEG analysis revealed significant differences between PTE and non-PTE groups, including higher delta frequency power and greater power variance in delta and theta frequencies in the PTE cohort.
• Predictive models combining qEEG and clinical features achieved an area under the curve of 0.76 (random forest).
• Logistic regression identified increased delta:theta power ratio and peak envelope as significant predictors of PTE risk.

Conclusions:

• Early phase qEEG features show potential in predicting the development of PTE in severe TBI patients.
• The developed predictive models can assist in identifying high-risk patients, guiding early clinical management, and informing patient selection for clinical trials.