The efficacy of virtual reality-enhanced EEG biofeedback in anxiety treatment: analysis of a digital health intervention

This study evaluates the therapeutic efficacy of Meditation therapy integrated with EEG biofeedback (MT-EB) for anxiety disorders and explores long short-term memory (LSTM) networks for personalizing meditation via EEG data. In a randomized controlled trial of 89 patients, home-based (H-EBM) and virtual reality (VR-EBM) approaches were delivered using portable EEG devices, VR glasses, and Android platforms. LSTM networks customized meditation scenarios, with outcomes measured by Hamilton Anxiety Scale (HAMA), Self-Rating Anxiety Scale (SAS), and biofeedback scores after three weeks. Post-treatment anxiety scores significantly decreased across all groups. While H-EBM + VR-EBM combination showed superior subjective SAS improvement (p < 0.05), HAMA reductions were comparable between groups. Biofeedback scores indicated general EEG improvements but failed to show VR-EBM superiority. Though LSTM-enabled personalization demonstrated feasibility, its advantage over manual protocols remained unproven. MT-EB treatment proved effective for anxiety, achieving outcomes comparable to cognitive behavioral therapy (CBT). The SAS score differentiation suggests VR-EBM integration may enhance subjective benefits, though biofeedback metrics did not confirm VR-specific advantages. LSTM applications for tailored interventions show promise but require further validation. These findings highlight MT-EB as a viable anxiety treatment, with perceptual framework innovations offering potential therapeutic strategies. Future studies should investigate long-term efficacy and broader implementation possibilities.