This article is drawn from one of my weekly "Neurofeedback & Chill" livestreams, where I run a short neurofeedback session on my own head while teaching a brain topic and taking questions. This week's topic was sensory and social processing. I've kept the audience questions in but stripped any names.
Why group sensory and social processing together?
These two systems share a lot of tissue, and they share most of the same approach when you want to regulate them. Here is the organizing principle I keep coming back to: the front of the brain is largely about the inside self, and the back of the brain is largely about the outside world. As you handle abstract, self-driven, interior material, that work happens forward of the brain's midpoint, into the frontal lobe. As you take in information and make meaning of it, that work happens behind the midpoint. Sensory and social processing live mostly behind the ears or further back, with a couple of exceptions I'll name.
When you map sensory and social difficulties, you see a recurring picture in the back-right quadrant. That's where I tend to look first.
What does disrupted sensory and social processing look like?
It shows up across a wide range. Some people get sensitivity to bright light or a nails-on-chalkboard reaction to certain sounds (chewing is a common irritability trigger). Others pick up too much social information at once: other people's anger, frustration, tone of voice. At the complex end, you get difficulty reading facial expressions and holding another person's perspective, the theory-of-mind work of treating someone else as a separate mind with their own thoughts and feelings. When that system runs too raw and wide open, you pick up too much and can't tease out the actual social signal.
I see dysregulation in sensory and social processing in anxiety, in some ADHD presentations, in autism, in concussion, and in post-viral conditions. Many roads lead to a flooded back-right quadrant.
How does the sensory system map onto the cortex?
Primary cortices take input straight from the outside world. Primary auditory cortex sits right above each ear. Primary visual cortex sits at the very back, at the tip of the occipital lobe.
The auditory system is one of the most expensive systems the brain runs. Sound starts as a compression wave in air, moves the eardrum, becomes a mechanical wave through the small bones, then a fluid wave down the cochlea that lands on hair cells based on frequency. The hair cells fire down the eighth cranial nerve into the brainstem, where the olivary nuclei compare both sides to localize sound, and then the signal projects to cortex with both same-side and opposite-side connections. Several transduction steps, an eight-way comparison system, and it never shuts off. Those bones keep moving while you sleep.
Because the auditory system runs continuously and pumps energy into the brain, anything that throws it off tends to show up clearly on a QEEG. It's common to see one auditory side running slow, with excess delta or theta, or flooded with beta because it can't filter the world, or running low alpha. When one side lags, the person experiences something like driving with a hard plastic tire on one wheel: you tap the gas, it spins, then catches. You start talking, they say "sorry, what?" If you say their name first and wait for the orienting response, then continue, they track you fine. Parents of spacey kids with ADD labels: make sure it isn't an auditory processing lag. You can test it by inserting a pause and watching whether the child catches up.
The visual system gets overwhelmed because primary visual cortex (V1) is also expensive. Close your eyes and it goes into alpha. Open them and it shifts toward beta as it processes the scene, unless you have ADD-type patterns and it stays in alpha. If you close your eyes and alpha does not rise, that's a marker of hypervigilance or anxiety: the visual system has learned to stay ready to process. After inflammation or concussion, that posterior tissue is often irritated, the world feels too bright and hard to filter, and you see disinhibited delta and theta.
The motor system makes a point worth holding onto. About 20 million cells leave the precentral gyrus to control the body. Of those, roughly 19 million stop in the cerebellum, and about 5 million continue down into the body. So around 95% of motor output is spent monitoring motor control, comparing intended fine and gross movement against what's actually happening. Even motor output is mostly sensory. Sensory processing involves an enormous amount of tissue.
Where do social cues get decoded?
Association cortices integrate input from other regions. Above that, the brain organizes around hub networks: the central executive network, the salience network, and the default mode network. The whole system connects to the gut and heart through the vagus nerve, which you can read through heart rate variability.
The right temporoparietal junction (TPJ) shows up more often than not when there are sensory and social difficulties. When people are flooded, things get loud, literally loud: other people's anger, their judgment, their eye contact. Next to the TPJ, the fusiform face area handles facial recognition.
We also monitor our own output to control it. Speech monitoring runs on both sides. On the right, you process prosody, the nonverbal rhythm and tone of speech. If you don't perceive prosody well, you don't monitor your own, and your voice goes monotone and often high-pitched. That's part of why many autistic speakers carry a flatter, higher-pitched delivery. On the left, Wernicke's area handles the meaning of language, with surrounding tissue making sense of incoming speech.
You see these same areas cramped up across very different conditions: brilliant and anxious people with sensory and social difficulty, autism, concussion (sports impacts and side-on car collisions are common with temporal-lobe involvement), and post-viral states. After COVID, it's common to see large blobs of delta, sometimes bilateral, sometimes one-sided, with the person describing brain fog and sensory irritability that looks a lot like a concussion. I don't know whether that comes from the metabolic hit, from old wear-and-tear interacting with new inflammation, or both. If you want to read more on the integration side, I cover it in depth in the sensory and social processing guide.
What protocol was I running, and why?
I ran C4 minus PZ with a low-beta reward, 4-7 Hz inhibit, 11.5-14.5 Hz reward, and 20-32 Hz inhibit. C4 sits over the right precentral gyrus, the supervisory system. I think of it as the passenger seat reading the map, telling you where to turn, reminding you to slow down. It uses a lot of low beta and supervises voluntary attention and physical inhibition (the ability to sit still). PZ is the posterior cingulate, part of the default mode network that scans and evaluates. Bracketing those two sites also picks up contributions from the right TPJ and surrounding tissue.
This pairing is not the protocol for sensory or social work for most people. If I had real social anxiety or sensory irritability, I'd work lower in the right quadrant, into the right TPJ directly. C4-PZ is a SMR-style setup that gave me a useful demonstration.
What I felt afterward: a body flush, a little focus activation from the C4 effect, a visual boost, and a bit of energy from raising beta (which may actually be fast alpha, since alpha runs up into the 10-13 Hz range). I started the session more relaxed than I ended it. Like a cup of coffee, smoother and clearer, with the visual field looking crisp and bright.
One practical note on signal quality. I fought low-amplitude signal for several minutes. Part of it was a virtual-ground issue in the software, part was an oily scalp slowing the paste, and one wire wasn't fully seated. Twenty-five years in and you still sometimes wrestle the signal. If you place electrodes, use different wire colors so you can sanity-check placement at a glance instead of tracing wires.
Do you need fancy games for neurofeedback to work?
Simple games work well. The reason matters. Neurofeedback works through implicit learning: the brain extracts rules from variable information through basic associative, operant-style conditioning. You can run this kind of conditioning on simple organisms, even single neurons. It doesn't require much cognitive horsepower, but it does require the brain to do statistical extraction, finding the pattern in the beeps and the swelling box and seeking more information. That's what we tap.
Here is why I avoid using movies and TV as the reward, with screen-dimming as the feedback. Social engagement floods or breaks implicit learning. The same tissue that extracts pattern and meaning also handles social meaning, social context, threat, and reward. Social processing is high-priority, so it takes over and makes it hard for the brain to work on the simple implicit information. I'm fairly confident that watching movies as your reward weakens neurofeedback. For years I thought certain software packages were just weaker; I now think the video reward itself reduces the training effect, either because it's social or because it's overstimulating enough to obscure the brain's math on simple stops and starts.
The exception is the person who will only sit still for a video: a deeply autistic child, a cognitively challenged adult, a teenager who won't engage otherwise. If a Shrek clip is the only way someone holds still while the brain trains, we run the Shrek clip. Partial training beats no training.
On reward schedules, the long-running debates about auto-thresholding versus fixed thresholds versus rapid thresholds mostly wash out. It all works as long as you give the brain enough information to do the heavy lifting. I auto-threshold every 30 seconds: move the criterion next to where the brain is, then wait for it to stay or move in the right direction.
How does neurofeedback help neuroinflammatory conditions like mold, Lyme, or long COVID?
Mold and Lyme look very similar on a brain map, and so do post-concussion and post-COVID states. They all tend to read as brain fog: slow waves, sluggish processing. Move carefully here. When someone is heavily inflamed and taxed, training can exceed their capacity. If a person feels hit by a truck after a session, that's overtraining, the same way you wouldn't keep climbing into a hyperbaric chamber if it left you feeling terrible. Let acute flares settle, sometimes just a few days, before pushing.
Target the metabolic side alongside the individual complaints. Neuroinflammatory states often bring tic-type phenomena, and after a concussion or fever you sometimes see tics or OCD-type eating patterns emerge from the cingulate and temporal lobes running hot. Work the three legs of the neurofeedback stool: sleep, stress, and attention. For people with this signature, I lean on a combination of EEG neurofeedback, HRV biofeedback, and HEG (hemoencephalography) for blood-flow training, plus heavier metabolic tools where resources allow: hyperbaric oxygen, red light therapy, and for some people cycling in and out of ketosis to build a strong anti-inflammatory effect. The metabolic hacks vary with the person. I care far more about carbohydrate intake in someone using diet to manage seizures than in someone adding muscle.
What about bipolar and schizophrenia?
You can't see bipolar or schizophrenia as a specific QEEG signature. The map looks off, usually some mix of fatigue and anxiety, sometimes front-midline and temporal-lobe features, sometimes sensory and social involvement. You don't read a thought disorder off the EEG. With auditory hallucinations, though, I've worked with more than a dozen people who all showed an auditory signature, usually on the right, alongside the intrusive experience.
Both respond to neurofeedback, but slower than I expect. Bipolar I tends to respond better than active schizophrenia. With several clearly schizophrenic clients I worked off and on for six months to a year, 100 to 150 sessions, and saw moderate gains in sleep, anxiety, executive function, and hygiene. During the work it often felt like something was fighting back. Then, a year or two or three later, I'd get a message and it was like talking to a twin who never had the condition: organized sentences, better boundaries, less pressured speech. That pressured quality is common in subacute schizophrenia and can tip into fear, aggression, or paranoia. Negative symptoms show as a wooden face, low movement, low expressiveness.
For anyone heavily dysregulated, sleep is the first target. Getting a strongly dysregulated person sleeping better will do 30 to 50% of the total work through knock-on effects. Even if you have no sleep complaint, watch how your sleep flexes after sessions and report it to your provider, because sleep shifts after neurofeedback almost always.
Which frequencies and sites relate to attachment?
Attachment is more subcortical than cortical. We know it's right-front because you get a strong release when you train up slow frequencies in the theta range at the right frontal pole, a very Sebern Fisher approach. My read is that we're reaching periamygdalar cortex (the amygdala itself has no EEG, but the cortical tissue around it does), the insula, and dorsolateral PFC tissue involved in avoidance and effort regulation. You can see how avoidance circuitry and relational-attachment circuitry would share resources: both scan for what isn't safe.
That right-front tissue is where I'd look for rejection sensitivity, ADHD-with-borderline splitting, rage, and reactive responses when boundaries get set. I call the F8/FT8 region the problem-child area. When you don't see a cortical alpha target there, you're likely working in subcortical tissue and have to lean on the protocol literature. If you want to do attachment and relational work seriously, study the people developing those protocols (Sebern Fisher, Ruth Lanius) rather than taking my framing as a recipe. I think rigorously about the brain and then mix in intuition once I'm looking at an individual's actual data, so a generic protocol never maps cleanly onto one person without testing.
Why is training theta tricky?
Excess resting theta in a chunk of brain is usually a failure mode: that tissue is stuck in high gear, disinhibited, sloppy in its regulation. So most protocols inhibit theta when it's excessive. Training theta up can reduce anxiety or amplify it powerfully, depending on where, in whom, and what their map looks like.
The deliberate exception is alpha-theta training, the Peniston-style approach, which trains up alpha and theta over visual and parietal association tissue until theta crosses above alpha and produces a hypnagogic, receptive state between waking and sleep. That's where visualization, insight, and emotional release happen. Useful, but powerful enough that I'd want a client reasonably resilient before going there. There are eyes-open variants that don't take you as deep, which can give a kind of dosing control for someone working with trauma who doesn't want to dissociate fully.
One caution from a viewer's experience: C4-PZ lifts mood and executive function for many people by speeding up alpha, but alpha resists being trained and tends to push back. About half the time, C4-PZ erodes sleep quality through that alpha manipulation. I often balance it with a segment of C3 or CZ minus A1 beta so I get the SMR benefit without disturbing alpha, which protects sleep.
The point of training on camera
I run the electrodes live to demystify the tool. People see it isn't complicated, and they hear the explanation as it happens. The goal is to get neurofeedback into the hands of anyone who wants it, as a usable tool rather than a black box.
If you want to see your own sensory and social tissue, a QEEG brain map shows where the right TPJ, the auditory regions, and the posterior cingulate are running. Peak Brain runs a yearly special with $250 off mapping at the offices in New York, St. Louis, Orange County, Los Angeles, Stockholm, and London, and remote mapping is available if you're not near one. From there, you build a protocol, test it, and iterate, the same way you'd adjust a workout based on how your body responds.
<<