Distinguishing Voluntarily Upregulation of Localized Central Alpha from Widespread Posterior Alpha.

Neurofeedback (NF) training based on alpha upregulation has been widely used on patient and healthy populations. However, active voluntary modulation of central or widespread posterior alpha in response to central alpha feedback is still ambiguous. The objective of this study is to confirm whether patients learn to truly increase alpha power and to determine if patients modulate central or widespread alpha power when alpha feedback is provided from central brain region. This EEG-based NF study was conducted on seven paraplegic patients with same injury type, pain location, and sensitization to ensure homogeneity. In addition to routine NF training sessions, various experiments were performed to compare alpha NF modulation received from C4 with alpha shift during cognitive tasks, occipital or parieto-occipital cortex, and Laplacian montage which is expected to separate localized alpha from widespread alpha, to attain objectives. Moreover, imaginary coherence analysis in alpha band was also performed to check whether C4 training site is coupled with other brain regions and to confirm whether activity at training site leads/lags the activity of other brain regions. The results indicate widespread alpha modulation in patients during regular NF sessions (p < 0.05) with large effect size (> 0.8), sufficiently high statistical power (> 80%), and a narrower confidence interval (CI) in response to NF provided from the central brain region reflecting less uncertainty and higher precision. However, small effect size obtained with Laplacian montage require patients to be trained with Laplacian feedback to achieve a reliable conclusion regarding localized alpha modulation. The outcomes of this study are not only limited to validate true alpha modulation in response to central alpha feedback but also to explore the mechanism of central alpha NF training.

Immediate effect of neurofeedback training on the pain matrix and cortical areas involved in processing neuropsychological functions.

OBJECTIVE: This study investigated the impact of neurofeedback training on the deeper cortical structures that comprise the "pain matrix" and are involved in processing neuropsychological functions. METHODS: Five paraplegic patients with central neuropathic pain received up to 40 sessions of neurofeedback training. They were asked to simultaneously modulate the relative power of the theta, alpha and beta bands, provided as a feedback from the sensorimotor cortex. The source localization technique was applied on EEG data recorded with 16 electrodes placed over the whole head. RESULTS: Neurofeedback training from the sensorimotor cortex induced effects on the pain matrix and in the areas involved in processing neuropsychological functions such as memory, executive functions and emotional regulations. Alpha and beta band activity was most increased in insular, cingulate and frontal cortex regions, and other areas corresponding to executive and emotional function processing. Theta band decreases were noted in the frontal, cingulate and motor cortices. In group analysis, theta and beta band activity was significantly reduced. CONCLUSION: The single channel electroencephalogram-based neurofeedback training produced effects on similar areas that are targeted in 19 channels standardized low-resolution brain electromagnetic tomography and expensive time-delayed functional magnetic resonance imaging feedback studies.

Study of change blindness EEG synchronisation using wavelet coherence analysis.

'Change blindness' is the failure to detect substantial changes in the visual field that normally would be noticed easily. This effect is of interest as by making comparisons with change detection, it may be informative of the brain activity that occurs during visual awareness. This paper proposes a wavelet coherence analysis for measuring the degree of synchronisation of multi-channel electroencephalograph (EEG) signals during experiments of change detection and change blindness in picture pairs presented between blank inter-stimulus intervals. The magnitude of wavelet coherence is found to exhibit significant differences between the hit and miss visual states. The ANOVA test verified the statistical significance of coherence differences (p less than 0.01) between hits (change detection) and misses (change blindness) in cross-channel synchronization.