Live Neurofeedback Session: Multi-Site SMR Protocol Design
In this week's neurofeedback and chill livestream, Dr. Hill demonstrated a live triple-site neurofeedback session while explaining his clinical approach to protocol design. The session focused on a C3-C4-CZ combinationâleft motor strip beta training, right motor strip SMR, and vertex SMRâproviding viewers with real-time insight into how experienced practitioners build and execute multi-site protocols.
Triple-Site Protocol: C3, C4, CZ Explained
Dr. Hill set up what he calls his preferred "triple protocol": C3 beta reward (left sensorimotor strip), C4 SMR training (right sensorimotor strip), and CZ SMR (vertex). This combination targets different aspects of brain regulation simultaneously.
The C3 beta training focuses on left hemisphere activation for cognitive processing, while the bilateral SMR components (C4 and CZ) work on thalamocortical stabilization. CZ at the vertex serves dual functionsâexecutive control and sleep enhancementâbecause it sits at the intersection of multiple cortical networks.
For the full deep dive on SMR mechanisms and clinical applications, see: SMR Neurofeedback: The Calm-Alert Brainwave.
Threshold Settings and Learning Parameters
Dr. Hill revealed his standard threshold configuration: 25-65-15 percentiles with 30-second auto-adjustments. These numbers determine how far from the brain's baseline the software sets reward criteria. The 25th percentile for inhibit thresholds means the targeted "bad" frequency must drop below where it sits 75% of the time. The 65th percentile for reward means the "good" frequency must exceed where it typically operates 35% of the time.
This creates what he describes as "adaptive pressure"âthe brain must trend in specific directions to keep the feedback game running. The 30-second updates prevent the system from becoming too easy or too difficult as the brain changes during the session.
Q&A Highlights: Clinical Problem-Solving
Question: Why do you recommend FC-PZ (frontal-posterior cingulate) training so frequently, especially with 7-10 Hz rewards?
Dr. Hill's Response: The anterior and posterior cingulate often get "stuck together" in problematic patterns. FC drives obsessive, perseverative thinkingâsongs stuck in your head, nail biting, OCD patterns. PZ generates more primitive threat responsesâfearfulness, rumination, trauma activation.
The FC-PZ protocol trains these regions to work independently rather than in rigid coupling. However, he often starts with PZ-A1 (posterior cingulate referenced to ear) first, since posterior cingulate hyperactivation tends to drive the whole system. Once that settles, the FC-PZ differential training becomes more effective.
Question: Do you ever see people who don't respond to neurofeedback, or whose EEGs move in the wrong direction?
Dr. Hill's Response: True non-responders are rareâmaybe one in a thousand cases where you see EEG changes but no experiential improvement. The brain data almost always moves in the trained direction. The bigger challenge is getting meaningful symptom change alongside the EEG shifts.
He emphasized that training purely to normalize brain maps without tracking subjective experience often produces "pretty brain scans" with limited real-world benefits. The key is balancing objective EEG changes with what the person actually experiences day-to-day.
Technical Setup Insights
During the live setup process, Dr. Hill demonstrated manual protocol configuration, explaining why most practitioners use pre-configured sessions. The "fiddly bits" of manual setupâelectrode placements, frequency bands, threshold percentages, session timingâcreate multiple failure points that can render training ineffective.
His approach uses differential electrode montages (C3-A1, C4-A2, CZ-A1) where the active sites are referenced to the ears. This removes common electrical noise and isolates the specific brain regions being trained.
Key Takeaways
- Multi-site protocols can target different regulatory systems simultaneouslyâcognitive activation, sensorimotor calming, and executive control
- Threshold settings (25-65-15) create adaptive learning pressure without overwhelming the brain's natural variability
- Electrode montages matter significantlyâdifferential referencing isolates target regions more effectively than single-site recordings
- Clinical intuition guides site selection, but tracking both EEG changes and subjective experience determines protocol success
- Sequential approaches often work better than complex multi-site protocols for beginnersâestablish one system before training multiple regions
The livestream highlighted how experienced neurofeedback practitioners think through protocol design in real-time, balancing theoretical knowledge with practical constraints and individual brain patterns.