Source-linked AI summary
Self-regulation of human brain activity using simultaneous real-time fMRI and EEG neurofeedback
Vadim Zotev, Raquel Phillips, Han Yuan, Masaya Misaki, Jerzy Bodurka
TL;DR
Existing neurofeedback methods typically use either EEG or rtfMRI, motivating a system that combines their complementary measurements in real time. The paper implements simultaneous multimodal rtfMRI-EEG neurofeedback and tests it during happy-memory emotional self-regulation. Healthy participants simultaneously regulated left-amygdala BOLD activity and frontal high-beta EEG asymmetry, demonstrating feasibility while highlighting substantial real-time EEG artifact challenges.
Problem
Neurofeedback commonly relies on either EEG or rtfMRI, while simultaneous EEG-fMRI creates an opportunity to combine electrophysiological and hemodynamic self-regulation.
Method
The study integrates simultaneous rtfMRI and EEG streams into real-time multimodal neurofeedback during happy autobiographical memory retrieval, targeting left-amygdala BOLD activity and frontal high-beta asymmetry.
Results
Participants simultaneously self-regulated hemodynamic activity in the left amygdala and electrophysiological activity measured by frontal high-beta EEG asymmetry.
Takeaways & Limitations
The proof-of-concept supports rtfMRI-EEG neurofeedback as a potential modality for cognitive neuroscience paradigms and cognitive therapeutic approaches, particularly for depression.
Takeaways & Limitations
Residual MR, cardioballistic, and muscle artifacts challenge real-time EEG neurofeedback, with residual MR and cardioballistic artifacts contributing about 50% of high-beta power after basic processing.
Abstract
from arXiv · showhide
Neurofeedback is a promising approach for non-invasive modulation of human brain activity with applications for treatment of mental disorders and enhancement of brain performance. Neurofeedback techniques are commonly based on either electroencephalography (EEG) or real-time functional magnetic resonance imaging (rtfMRI). Advances in simultaneous EEG-fMRI have made it possible to combine the two approaches. Here we report the first implementation of simultaneous multimodal rtfMRI and EEG neurofeedback (rtfMRI-EEG-nf). It is based on a novel system for real-time integration of simultaneous rtfMRI and EEG data streams. We applied the rtfMRI-EEG-nf to training of emotional self-regulation in healthy subjects performing a positive emotion induction task based on retrieval of happy autobiographical memories. The participants were able to simultaneously regulate their BOLD fMRI activation in the left amygdala and frontal EEG power asymmetry in the high-beta band using the rtfMRI-EEG-nf. Our proof-of-concept results demonstrate the feasibility of simultaneous self-regulation of both hemodynamic (rtfMRI) and electrophysiological (EEG) activity of the human brain. They suggest potential applications of rtfMRI-EEG-nf in the development of novel cognitive neuroscience research paradigms and enhanced cognitive therapeutic approaches for major neuropsychiatric disorders, particularly depression.
Introduction
The paper introduces the first simultaneous multimodal rtfMRI-EEG neurofeedback system, integrating hemodynamic and electrophysiological signals in real time. It applies this proof-of-concept system to emotional self-regulation using happy autobiographical memories, targeting left-amygdala BOLD activity and frontal high-beta EEG asymmetry.
- Simultaneous EEG-fMRI creates opportunities to examine electrophysiological correlates of rtfMRI neurofeedback, validate EEG neurofeedback with fMRI, and adapt feedback dynamically.
- The study reports the first implementation of simultaneous multimodal rtfMRI-EEG neurofeedback for training emotional self-regulation.
- The novel system integrates simultaneous rtfMRI and EEG data streams into a multimodal neurofeedback display for brain neuromodulation.
- The rtfMRI signal targeted BOLD activation in the left amygdala, while EEG feedback represented frontal F3–F4 power asymmetry in the high-beta band.
- High-beta asymmetry was selected instead of alpha partly because EEG-fMRI artifacts are substantially reduced in that band and because high-beta activity is relevant to depression.
- Healthy participants retrieved happy autobiographical memories while trying to raise simultaneous rtfMRI-nf and EEG-nf bars.
Data acquisition
The study acquired fMRI and EEG simultaneously in an MRI-compatible setup and processed both streams in real time. fMRI feedback was updated every 2 seconds, while EEG asymmetry processing used artifact-corrected F3 and F4 signals.
- Functional and structural MR images were acquired with a 3 T MRI scanner and an 8-channel receive-only head coil array.
- The fMRI acquisition used a single-shot gradient-recalled EPI sequence with SENSE and the reported spatial, timing, and scan-duration parameters.
- EEG was recorded simultaneously using an MR-compatible 32-channel cap with electrodes arranged according to the international 10-20 system.
- Real-time AFNI processing registered EPI images and exported mean left-amygdala ROI signals for neurofeedback.
- The rtfMRI neurofeedback signal was computed as percent change from the preceding 40-second Rest block and updated every 2 seconds.
- Real-time EEG processing partially removed MR and cardioballistic artifacts before computing F3 and F4 power spectra and high-beta asymmetry.
- Offline fMRI preprocessing included cardiorespiratory artifact correction with RETROICOR, slice-timing correction, and volume registration.
EEG data analysis
The study processed simultaneous EEG-fMRI data to estimate high-beta frontal asymmetry and examine its relationship with fMRI responses during neurofeedback. Results support simultaneous regulation while highlighting persistent real-time artifact challenges.
- EEG preprocessing: EEG data were corrected for MR and cardioballistic artifacts, downsampled to 250 S/s, filtered, and motion-affected intervals were excluded.Offline processing additionally used ICA to remove residual MR, cardioballistic, muscle, and eye-movement artifacts.
- Neurofeedback results: Participants increased average left-amygdala BOLD activation and frontal high-beta EEG asymmetry across neurofeedback and transfer runs, with significance in selected runs.EEG asymmetry was significant in Run 3 (t(5)=2.75, p<0.04), while left-amygdala activation was significant in the Practice run (t(5)=4.81, p<0.005).
- Artifact effects: Removing additional artifacts generally made positive Happy Memories-versus-Rest asymmetry changes more pronounced, indicating artifacts did not fully explain them.Residual MR and cardioballistic artifacts contributed about 50% of average high-beta power after basic processing, while muscle artifacts contributed about 20%.
- EEG-informed fMRI analysis: EEG-informed PPI analysis found stronger high-beta asymmetry-related fMRI interactions during Happy than Count conditions, including the left amygdala target.The authors interpret this as evidence that the EEG- and rtfMRI-based neurofeedback signals were mutually compatible.
- Implications: Simultaneous rtfMRI-EEG neurofeedback may complement slower BOLD feedback with faster EEG feedback and support real-time protocol or strategy adjustments.The authors also propose that EEG feedback could eventually approximate some rtfMRI feedback effects, although this remains a future application.
Conclusion
The study demonstrates the feasibility of simultaneously self-regulating hemodynamic and electrophysiological brain activity using multimodal rtfMRI and EEG neurofeedback.
- The first implementation of simultaneous multimodal rtfMRI and EEG neurofeedback enabled simultaneous self-regulation of hemodynamic and electrophysiological brain activity.