Neurophysiological signatures of mild traumatic brain injury in the acute and subacute phase

Abstract

Background: Mild traumatic brain injury (mTBI) affects 48 million people annually, with up to 30% experiencing long-term complaints such as fatigue, blurred vision, and poor concentration. Assessing neurophysiological features related to visual attention and outcome measures aids in understanding clinical symptoms and prognostication.

Methods: We recorded EEG and eye movements in mTBI patients during a computerized task performed in the acute (< 24 h, TBI-A) and subacute phase (4-6 weeks thereafter). We estimated the posterior dominant rhythm, reaction times (RTs), fixation duration, and event-related potentials (ERPs). Clinical outcome measures were assessed using the Head Injury Symptom Checklist (HISC) and the Extended Glasgow Outcome Scale (GOSE) at 6 months post-injury. Similar analyses were performed in an age-matched control group (measured once). Linear mixed effect modeling was used to examine group differences and temporal changes within the mTBI group.

Results: Twenty-nine patients were included in the acute phase, 30 in the subacute phase, and 19 controls. RTs and fixation duration were longer in mTBI patients compared to controls (p < 0.05), but not between TBI-A and TBI-S (p < 0.05). The frequency of the posterior dominant rhythm was significantly slower in TBI-A (0.6 Hz, p < 0.05) than TBI-S. ERP mean amplitude was significantly lower in mTBI patients than in controls. Neurophysiological features did not significantly relate to clinical outcome measures.

Conclusion: mTBI patients demonstrate impaired processing speed and stimulus evaluation compared to controls, persisting up to 6 weeks after injury. Neurophysiological features in mTBI can assist in determining the extent and temporal progression of recovery.

Keywords: EEG; ERP; Eye tracking; Mild traumatic brain injury; Visual attention.

Fig. 1

Reaction times. Total RT and RT subcomponents from the CRT task. Total RT is higher for TBI-A compared to TBI-S and controls. Visual reaction time (VRT) is higher in TBI-A, as well (mean response change from controls to mTBI is 45 ms). Processing speed (PS) is comparable between TBI-A and TBI-S (804 vs 806 ms, respectively), but was notably lower in controls (714 ms). Fixed-effect coefficients for the mTBI term and post hoc analysis between controls and mTBI singularly are significant (p < 0.05) for total RT, VRT (TBI-A, only), and PS, but not for saccadic latency (SL)

Fig. 2

Fixation duration. Fixation duration, defined as the sampling time of the eyes within the area of interest around the target stimulus during the CRT task. Fixation duration is significantly impacted by mTBI. Post hoc analysis determined discrepancies between controls and TBI-A or TBI-S singularly. The mean response change from controls to mTBI is 139 ms

Fig. 3

PDR peak frequency. Peak frequency of parieto-occipital channels in the alpha band during resting EEG with eyes closed. A mean difference of 0.6 Hz is found between TBI-A and TBI-S (p = 0.02). Red line: median; diamond: mean

Fig. 4

Event-related potentials. Top: grand average ERP for channel P3, Pz, and P4 of the three groups. The colored area around the averaged signal represents the standard error of the mean for each time point. Bottom left: the mean amplitude between 250 and 750 ms is significantly different across groups. The mean difference (MD) between controls and both TBI groups is 3.4 µV. No significant difference is found between TBI-A and TBI-S. Bottom right: peak latency between 250 and 750 ms. No significant difference across groups is found

Fig. 5

GOSE outcome and neurophysiological metrics. Left: association between GOSE outcomes and total RT for TBI-A and TBI-S. Both mTBI patients with incomplete (pink) and complete (purple) recovery in the acute and subacute phases show a similar median RT (median Inc Rec TBI-A and TBI-S = 1.2 s; median Com Rec TBI-A = 1.3 s, TBI-S = 1.2 s; Spearman’s correlation RT vs GOSE: TBI-A: p = 0.1; TBI-S: p = 0.4). Right: association between GOSE outcomes and peak frequency in the posterior, alpha band during resting EEG with eye closed for TBI-A and TBI-S. Both groups show a similar median peak frequency with complete (purple) and incomplete (pink) recovery of GOSE scores (median Inc Rec TBI-A = 9.3 Hz, TBI-S = 9.5 Hz; median Com Rec TBI-A = 9.8 Hz, TBI-S = 9.4 Hz; Spearman’s correlation peak frequency vs GOSE: TBI-A: p = 0.2; TBI-S: p = 0.6). Inc Rec, incomplete Recovery; Com Rec, complete recovery