Disclaimer: The content below was generated with the assistance of AI and then reviewed and edited by BrainMaster Technologies, Inc. It is provided for educational and informational purposes only and does not constitute medical advice.
Overview #
Making Coherence Coherent: Brain Connectivity Assessment and Training (Collura, 2007) provides a foundational technical explanation of EEG connectivity, coherence metrics, and their use in assessment and neurofeedback training. The document outlines the mathematical and practical basis for several connectivity measures, with an emphasis on BrainMaster implementations and real-time training applications.
Purpose of Connectivity Training #
Connectivity training aims to evaluate and influence how different brain regions communicate. Key goals include:
-
Reflecting whole-brain functional relationships
-
Measuring information sharing between sites
-
Assessing speed and stability of inter-site interactions (phase, amplitude, timing)
-
Enabling real-time feedback for neurofeedback training
Theoretical Framework #
Generalized Connectivity Model #
The model (page 3) illustrates neuronal assemblies exchanging information with varying degrees of coupling, coherence, and phase timing. It serves as the conceptual underpinning for all connectivity metrics.
System Identification & Parameter Estimation #
Described on page 4, this framework defines how real neural properties are converted into measurable EEG parameters through assumptions, definitions, and algorithmic estimations.
Connectivity Measures Explained #
1. Pure Coherence #
-
Measures phase stability between signals
-
Ignores amplitude differences and absolute phase
-
Indicates shared information and timing stability
-
Calculated via FFT or quadrature filters
A comparison study on page 10 demonstrates strong concordance between BrainMaster coherence and NeuroGuide coherence estimates.
2. Training Coherence (Similarity) #
-
Sensitive to phase and amplitude matching
-
Ideal for synchrony-based training
-
Uses quadrature filters for real-time analysis
3. Spectral Correlation Coefficient (SCC) #
-
Measures similarity of spectral amplitude shapes
-
Phase-independent
-
Based on FFT amplitude domain
-
Can show high similarity even with random but similarly shaped spectra
4. Comodulation #
-
Correlation between amplitude envelopes over time
-
Phase-independent
-
Reflects how channels “wax and wane” together
5. Phase Metrics #
-
Evaluate relative timing and phase separation
-
Provide insights into speed of information sharing
-
Useful in synchrony-training protocols
6. Sum and Difference Channels #
Sum Channel #
-
Enhances in-phase relationships
-
Sensitive to synchrony
-
Useful for uptraining coherence
Difference Channel #
-
Emphasizes divergence (similar signals cancel)
-
Useful for downtraining coherence
Advanced BrainMaster Connectivity Tools #
BrainScape Joint Time-Frequency Analysis (JTFA) #
Pages 23–26 show visualizations of recombined channels (e.g., F3/F4, C3/C4), highlighting lateralized and regional communication patterns.
Z-Score Connectivity #
Available Z-Scores #
-
Absolute & relative power
-
Ratios
-
Asymmetry
-
Coherence
-
Phase
Based on a normative database of >600 subjects, age-adjusted for EO/EC conditions.
Real-Time Z-Score Training #
-
2-channel: 76 targets
-
4-channel: 248 targets
Shown on pages 28–29.
Z-Score Training Methods #
-
Directional up/down training
-
Range-based training via Rng()
-
PercentZOK() training, rewarding when a percentage of Z-Scores fall within a target range
Figures on pages 36–38 demonstrate the impact of adjusting thresholds and target-window widths.
Summary of Key Insights #
-
Multiple connectivity metrics exist, each offering unique advantages for assessment and neurofeedback.
-
Understanding methodological differences is critical for proper clinical interpretation.
-
Normative Z-score databases significantly enhance interpretability and provide structured frameworks for training.
-
BrainMaster systems provide real-time implementations of these methods, enabling flexible training strategies.
