Identification of Concussion Subtypes Based on Intrinsic Brain Activity

Area of Science:

• Neuroscience
• Neurology
• Biomedical Engineering

Background:

• Concussion pathophysiology remains incompletely understood, hindering effective treatment.
• Identifying distinct patient subgroups at injury onset could improve prognosis and care.
• Resting-state quantitative electroencephalography (qEEG) offers a potential tool for objective brain state assessment.

Purpose of the Study:

• To investigate the existence of intrinsic brain activity-based concussion subtypes.
• To define these subtypes using distinct resting-state qEEG profiles at the time of injury.
• To explore the relationship between identified subtypes and clinical outcomes, such as return to activity.

Main Methods:

• Retrospective multicenter cohort study including 771 participants (aged 13-70) with mild head injuries.
• Acquisition of 2-minute artifact-free resting-state electroencephalography (EEG) within 120 hours of injury.
• Application of bootstrapped coclustering algorithm to 471 qEEG features for subtype identification and validation.

Main Results:

• Five distinct concussion subtypes were identified based on qEEG power, connectivity, and complexity profiles.
• Subtype membership demonstrated a statistically significant association with the time to return to activity.
• The clustering algorithm achieved high agreement (98.87%) in defining disjoint subtypes.

Conclusions:

• Distinct physiological subtypes exist within the concussed population, identifiable via resting-state qEEG.
• These subtypes suggest varying underlying pathophysiological mechanisms.
• Identification of qEEG-based concussion subtypes holds potential for personalized prognosis and optimized treatment strategies.