Author(s): Zijp, Timon van (2024)
Abstract:
Mild traumatic brain injury (mTBI) is common (incidence of 600+ per 100,000 per year). Most patients recover well within months, 15 – 30% suffer complaints long after trauma. No reliable method exists to early predict outcome in mTBI patients. In this explorative study, we searched for possible correlations between EEG characteristics and outcome. We used EEG-data from 23 acute (<24h after trauma) and 26 subacute (4 – 6 weeks after trauma) patients (23 overlap). After 6 months, outcome was assessed with the Head Injury Symptom Checklist and Glasgow Outcome Scale – Extended questionnaires. We performed four different signal analysis techniques: relative power, power variability, power symmetry and coherence. We found three possible correlations. Patients with fewer persistent complaints have lower relative alpha power in the acute phase compared to the subacute phase (correlation value = -0.44, P = 0.04). Subacute patients with more persistent complaints have lower theta power variability than healthy average (correlation value = -0.43, P = 0.03). Acute patients have a lower coherence than healthy average; those with a poor recovery have a lower coherence than those with good recovery (P = 0.04). In future research we recommend validating our findings or implementing our findings in a machine learning system.
Document(s):
VanZijp_MA_TNW.pdf