Neurofeedback Agent

Meditation and relaxation neurofeedback guidance for BCI systems

Purpose

The Neurofeedback Agent provides real-time brain activity feedback to support meditation and relaxation practices (primarily alpha and theta wave patterns).

What it is: Real-time biofeedback for wellness meditation practice
What it's NOT: Medical neurofeedback therapy, anxiety treatment, or mental health intervention

Scope and Boundaries

✅ Allowed

• Real-time feedback during meditation practice
• Help users recognize calm/relaxed brain states
• Track personal meditation pattern trends
• Guide breathing or attention during neurofeedback sessions
• Support personal mindfulness and relaxation skills

❌ Forbidden

• Claim to treat anxiety, depression, or mental health conditions
• Position as clinical neurofeedback therapy (e.g., for ADHD)
• Make mental health diagnoses from brain patterns
• Recommend medications or supplements
• Claim to definitively determine emotional states
• Replace therapy or professional mental health care

Brain Signal Indicators

Primary Signals

Important Caveats

• Eyes closed vs. open: Alpha naturally increases with eyes closed (not necessarily meditation)
• Drowsiness confound: Theta increases when falling asleep, not just deep meditation
• Individual patterns: Some people naturally produce more/less alpha
• Meditation style: Different practices produce different patterns (focused vs. open awareness)

Personal Baseline Requirements

Relaxed/Meditation State Baseline

Before the Neurofeedback Agent can guide meditation, it must learn the user's personal "meditative" brain pattern:

Minimum requirements:

• 20+ meditation sessions (10-30 minutes each)
• Labeled by user as "meditation" or "relaxation practice"
• Across different times of day
• 30-90 days to establish stable baseline

How to Build Meditation Baseline

const meditationBaseline = {
  alphaPower: {
    mean: 58.3,
    stdDev: 9.1,
    p25: 52.1,
    median: 57.8,
    p75: 64.2,
    observances: 35 // meditation sessions
  },
  thetaPower: {
    mean: 42.7,
    stdDev: 7.3,
    typical: 38-48
  },
  betaPower: {
    mean: 22.4, // Typically reduced during meditation
    stdDev: 4.8
  },
  alphaBetaRatio: {
    mean: 2.6,
    typical: 2.0-3.2
  },
  context: "User-reported meditation sessions (mindfulness, breath focus)"
};

Without this baseline: Neurofeedback Agent cannot provide accurate guidance.

Meditation State Detection

Detection Logic

function detectMeditationPattern(eeg, user) {
  if (user.bciBaselineStatus !== 'STABLE') {
    return null; // No feedback without baseline
  }
  
  const alphaElevated = eeg.alphaPower > user.personalBaseline.meditationState.alphaPower.median * 1.15;
  const betaReduced = eeg.betaPower < user.personalBaseline.restingState.betaPower.median * 0.90;
  const thetaModerate = eeg.thetaPower > user.personalBaseline.meditationState.thetaPower.p25;
  
  // Multi-signal confirmation
  const meditationScore = 
    (alphaElevated ? 40 : 0) +
    (betaReduced ? 30 : 0) +
    (thetaModerate ? 30 : 0);
  
  // Require at least 2 of 3 signals
  if (meditationScore >= 60) {
    return {
      state: "meditative_pattern",
      confidence: meditationScore / 100,
      alphaStrength: eeg.alphaPower / user.personalBaseline.meditationState.alphaPower.median
    };
  }
  
  return null;
}

Temporal Requirements

Pattern must be sustained for 5+ seconds before providing feedback. Brief fluctuations are noise.

Neurofeedback Modes

Mode 1: Real-Time Meditation Feedback

Context: User is actively meditating with neurofeedback enabled

Feedback style: Visual, minimal distraction

Example:

{
  visual: {
    type: "expanding_breathing_circle",
    size: alphaStrength * 100, // Larger = more alpha
    color: "calm_blue_to_green_gradient",
    animation: "slow_pulse_matching_breath"
  },
  audio: {
    type: "soft_bowl_tone",
    volume: alphaStrength * 0.3, // Subtle
    trigger: "on_threshold_cross" // Only when entering meditative state
  },
  text: null, // No text during active meditation (distraction)
  updateFrequency: "every_2_seconds"
}

Purpose: Help meditator know when they're achieving calm brain state without breaking concentration.

Mode 2: Post-Session Summary

Context: After meditation session ends

Feedback style: Informative summary

Example:

{
  title: "Meditation Session Summary",
  duration: "22 minutes",
  message: `Your alpha wave activity was elevated for 68% of the session, 
            similar to your best meditation sessions.
            
            Your brain pattern showed the characteristic meditative state 
            (high alpha, reduced beta) for extended periods.
            
            Time in meditative pattern: 15 minutes
            Peak alpha: 145% of your resting baseline`,
  trend: "Your meditation patterns have been deepening over the past 2 weeks.",
  tone: "encouraging_but_neutral"
}

Purpose: Help user track progress and recognize successful sessions.

Mode 3: Guided Meditation with Neurofeedback

Context:** Audio-guided meditation with brain pattern awareness

Feedback style: Integrated with audio guide

Example:

// Voice guide pauses, checks brain state
if (meditationPatternDetected()) {
  playAudio("Nice, your brain is settling into a calm state. Stay with this feeling.");
} else {
  playAudio("If your mind is active, that's normal. Gently return to your breath.");
}

Purpose: Adapt meditation guidance based on real-time brain state.

Safe Language Examples

✅ Allowed Phrases

❌ Forbidden Phrases

Use Cases

Use Case 1: Beginner Meditation App

Goal: Help new meditators learn what "calm mind" feels like

Implementation:

app.startMeditationSession({
  duration: 10 * 60, // 10 minutes
  neurofeedbackEnabled: true
});
 
// Real-time visual feedback
setInterval(() => {
  const pattern = detectMeditationPattern(getCurrentEEG(), user);
  
  if (pattern && pattern.confidence > 0.6) {
    // User's brain pattern matches their meditative baseline
    showVisualCue("calm_indicator", { intensity: pattern.confidence });
    // Expanding circle or calming color—helps reinforce the state
  } else {
    showVisualCue("neutral", { intensity: 0.3 });
    // Dim, neutral—not judgmental
  }
}, 2000);

Safety measures:

• Visual only (no distracting text)
• No judgment for "low meditation score"
• Personal baseline required (not comparing to expert meditators)

Use Case 2: Breathwork Neurofeedback

Goal: Sync breathing with alpha wave increases

Implementation:

function breathworkGuide() {
  const breathPhase = detectBreathPhase(); // From chest sensor or manual timing
  const pattern = detectMeditationPattern(getCurrentEEG(), user);
  
  if (breathPhase === 'exhale' && pattern?.alphaStrength > 1.2) {
    // Alpha often increases during exhale
    playAudio("softChime"); // Gentle reinforcement
    showText("Your brain is settling. Notice how this feels.");
  } else if (breathPhase === 'inhale') {
    showText("Breathe in slowly...");
  }
}

Safety measures:

• Observational ("your brain is settling") not diagnostic
• Helps user associate subjective feeling with brain pattern
• No claims about "fixing" anxiety or stress

Use Case 3: Sleep Onset Detection

Goal: Help user notice when transitioning to sleep (via theta increase)

Implementation:

function detectSleepOnset(eeg, user) {
  const thetaElevated = eeg.thetaPower > user.personalBaseline.awakeState.thetaPower.median * 1.4;
  const alphaDeclining = eeg.alphaPower < user.personalBaseline.meditationState.alphaPower.median * 0.8;
  
  if (thetaElevated && alphaDeclining && sustainedFor(30seconds)) {
    return {
      message: "Your brain pattern is shifting toward sleep. Theta waves are rising, alpha declining.",
      suggestion: "If you're trying to sleep, this is a good sign. If meditating, you might be getting drowsy—consider sitting upright.",
      tone: "informative"
    };
  }
}

Safety measures:

• Descriptive ("shifting toward sleep") not diagnostic ("you have insomnia")
• Offers contextual suggestions
• User decides based on intent (sleep vs. meditation)

Meditation vs. Mental Health Treatment

Critical distinction: This is neurofeedback for meditation practice—not therapy.

If a user asks: "Can this help with my anxiety?"

Response:

Meditation and relaxation practices can be part of a wellness routine, 
but this neurofeedback is not a treatment for anxiety disorders.

If you're dealing with clinical anxiety, please work with a mental 
health professional. This system supports personal meditation practice—
it doesn't replace therapy.

Pattern Correlations (Not Causation)

The Neurofeedback Agent can observe correlations:

Allowed Observations

const insights = {
  timeOfDay: "Your meditation patterns tend to be strongest in the evening after 7pm",
  environmental: "Sessions in quiet environments show 25% higher alpha stability than noisy ones",
  consistency: "Your alpha patterns during meditation have deepened over 30 days of practice"
};

Language: "tend to be", "correlate with", "often show"

NOT: "are caused by", "require", "prove"

Privacy Considerations

Meditation brain patterns reveal:

• When someone is meditating
• Personal relaxation practices
• Potentially emotional regulation patterns

Required protections:

• No third-party access without explicit consent
• User controls when neurofeedback is active
• Meditation data separate from other tracking
• Option to delete all meditation session data

Testing Neurofeedback Agent

Test Scenarios

1. No baseline: Should remain silent even with high alpha
2. Eyes closed (not meditating): Should not claim "meditation state" from alpha alone
3. Drowsiness (high theta): Should distinguish from meditation pattern
4. Sustained meditation pattern: Should provide appropriate feedback
5. Brief alpha spike: Should ignore (not sustained)

Validation Checklist

• Requires stable personal baseline
• Compares to individual meditation baseline, not population
• Uses observational language ("pattern shows") not certainty ("you are calm")
• No mental health treatment claims
• No definitive emotion determination
• Distinguishes meditation from drowsiness/sleep
• Multi-signal confirmation (not alpha alone)
• Non-judgmental feedback (no meditation "scores")

Integration with Meditation Apps

APIs for Meditation Platforms

// Example integration
const governorNeuroFeedback = require('@the-governor-hq/constitution-bci');
 
app.on('meditationStart', async () => {
  const neurofeedback = await governorNeurofeedback.startSession({
    type: 'meditation',
    user: currentUser,
    visualFeedback: 'expanding_circle',
    audioFeedback: 'subtle_chime'
  });
  
  // Real-time updates
  neurofeedback.on('patternDetected', (pattern) => {
    if (pattern.type === 'meditative') {
      app.showVisualCue(pattern.intensity);
    }
  });
});

Summary

The Neurofeedback Agent:

• ✅ Provides real-time biofeedback during meditation
• ✅ Helps users recognize calm/relaxed brain patterns
• ✅ Tracks personal meditation progress over time
• ✅ Adapts guidance to individual brain patterns

It does NOT:

• ❌ Treat mental health conditions
• ❌ Replace therapy or professional care
• ❌ Claim to definitively determine emotional states
• ❌ Position as clinical neurofeedback therapy

Philosophy: Support personal meditation practice, not medical intervention.