Efficacy of AI-Based Wearable EEG Devices in Detecting Psychological Distress and Monitoring Relaxation Interventions in Active-Duty Armed Personnel.

Abstract

High-stress workplaces, such as active-duty armed forces, make objective, continuous assessment of distress difficult.Smart wearables embedded with AI would provide an ideal alternative to self-reporting.To assess the ability of a commercially available AI-based EEG device, to compare the effectiveness of these techniques, and to compare their effectiveness on stressrelated neurophysiological biomarkers over the duration of a 12-session protocol for armed staff.A withinsubject repeated-measures design was employed with 200 active-duty armed personnel (mean age= 34.2 years, SD= 8.7).Participants completed 12 structured experimental sessions and incorporated five relaxation techniques, which are pranayama, dhyana, nadayoga, visual meditation, and mantra chanting.EEG band recordings measured pre-and post-session measured alpha (8-13 Hz), and beta (13-30 Hz) band power, which were measured as primary biomarkers of psychological arousal.The Depression, Anxiety, and Stress Scale-14 (DASS-21) was administered at baseline and follow-up.Paired-samples t-tests and repeated-measures ANOVA were used for multiple comparisons.All five interventions showed statistically significant pre-to-post changes in both Alpha power and Beta power (all p< 0.001).Alpha power increased significantly (M_pre = 10.03Hz, M_post = 10.34Hz, Δ = +0.31Hz, t (199) = -12.02,p < 0.001, Cohen's d = 0.85; 95% CI [0.67, 1.03]), indicating cortical relaxation.Beta power decreased significantly (M_pre = 20.01Hz, M_post = 17.30Hz, Δ = -2.71Hz, t (199) = 56.46,p < 0.001, Cohen's d = 3.99; 95% CI [3.75, 4.23]), indicating reduced cortical arousal.When comparing efficacy, mantra chanting produced the largest beta reduction (Δ = -0.36Hz each, d = 1.28), and pranayama and dhyana produced the largest alpha augmentation (Δ = +0.36Hz each, d= 0.92 and d= 0.91, respectively).Total DASS-21 scores decreased 54.4% from baseline (M= 16.74, SD= 8.23) to post-protocol (M= 7.62, SD= 4.15), t (199) = 15.37,p< 0.001, d= 1.09; 95% CI [0.91, 1.27], suggesting a clinically meaningful change in psychological distress.This is the strongest empirical evidence to date of an objectively measurable neurophysiological signature of reduced psychological distress from AI-enabled wearable EEG devices in armed personnel.When objectively measurable biomarkers of reduced neurophysiological stress converge with reduced psychological distress, this offers converging evidence of the neural correlates of these reductions and, in this way, confirms the neurophysiological meaningfulness of the reduction and, in such a context, can form the robust basis for the deployment of wearable EEG platforms into military occupational health surveillance systems as stigmareducing, deployable alternatives to subjective-only assessment paradigms.