Evaluation of PainWaive: A consumer-grade EEG headset for remotely delivered neu-rofeedback and monitoring in chronic pain

Affordable home-based electroencephalography (EEG) headsets could widen access to EEG assessment, but require rigorous validation before research or clinical use. Here, we evaluated a custom-developed 2-channel sensorimotor headset (PainWaive) intended for remote neuro-feedback and longitudinal monitoring in chronic pain. Eighty participants (47 female; mean age 24.0 years, SD 7.9) completed two resting-state sessions with PainWaive and a research-grade 64-channel EEG system (LiveAmp), under eyes-open (EO) and eyes-closed (EC) conditions. Alpha, beta and theta power and peak alpha frequency (PAF) were derived from homologous sensorimotor channels (C1/C2). Relative reliability was quantified with intraclass correlation coefficients (ICCs), absolute reliability with SEM%, and cross-device consistency with between-device ICCs and Pearson correlations of overall spectral shape. ICCs/correlations were interpreted using pre-specified thresholds: fair 0.20-0.39, moderate 0.40-0.59, good 0.60-0.79, excellent [≥]0.80. PainWaive and LiveAmp showed comparable absolute reliability across metrics (similar SEM%). Under EC, PainWaive reliability was excellent for alpha (0.81), theta (0.85) and PAF (0.94), and good for beta (0.72). Under EO, reliability was excellent for alpha (0.82), good for beta and PAF (0.61-0.72), and moderate for theta (0.59). Spectral-shape correlations between devices were excellent (r>0.90). Cross-device ICCs were good under EC for alpha/theta/PAF (ICC=0.66-0.77) though fair for beta (0.35). Under EO, ICCs were good for alpha (0.62), moderate for PAF (0.53), and fair for beta/theta (0.26-0.32). To assess performance under real-world use, we additionally analysed 2 clinical samples of individuals (total n = 8) with chronic pain who each completed 20 home-based neurofeedback sessions using PainWaive (160 sessions total). Within-session stability was good-to-excellent across metrics (ICCs>0.72). Overall, our findings suggest PainWaive is a reliable tool for the assessment of EEG metrics, supporting its use in research and clinical applications.