🎙️ Join us for a Live NeuroNoodle Neurofeedback Podcast Q&A! 🧠✨

uh 10 million 12 million you know so the it's it's definitely an economic disaster and we'll see what happens in La over the next couple years when things get rebuilt out we'll see well I'm just happy Dr Hill that you are there for selfish reasons because you are always here at Q&A how was your Q&A on Monday it was pretty good what is the one that you do on Monday so every Monday on YouTube YouTube do Hill drill um I do a neura feedback and chill at 6 PM Pacific or 9 900 P p.m. Eastern and hang out for about an hour and usually rant on some particular topic that I think we should you know have more perspective on and take more control over some physiology focused biohacking and typically start that with some neuro feedback and I explain a particular protocol that I'm doing or I show the process of neuro feedback in real time so folks understand that it's actually not that hard to do neuro feedback it's kind of hard to know what to do in the process the training process that we use but the techniques were not that complicated to set up and execute in a neur feedback piece of software so I show people that on Mondays yeah and it all starts with the EG right can't do anything without that the raw data it does um no matter the provider no matter the style or the software uh all neuro feedback is training something inside the brain you know that's the definition training signals inside the central nervous system and most neur feedback is EEG um not every EEG neuro feedback provider will use the EEG in an assessment context and not every assessment is created equal so we have these things called qegs which are full head uh resting baselines that are uh phenotypes or biomarkers or patterns that we tend to you know trait markers we carry around and they're not perfect this is the problem in neuro feedback and in qeg and in brain mapping is the qeg is valid in terms of like it's stable data if I did a qeg or an EEG on you today and one in a month one in a month after that it would be the same EG roughly at a high level but then qeg takes that and Compares it to the population averages right so we look at your brain and what sticks out and is unusual but guess what people are weird so good job be weird the the point's not necessarily to judge difference from average as a problem when it comes to the EEG it's to use the ways in which things stick out and uh uh how things are unusual to inform a perspective and then we also use performance testing like ADHD testing alongside that to to get out hey we have a question go hia hi Haida H I like it so what is the arousal model in neuro feedback it's one of the um I would I would kind of consider it one of the core like schools of thought in neuro feedback in some ways the rasal model the idea that we have a mix of uh brain waves that affect like how active its bit of tissue is and so you can think about different parts of the cortex will make a combination of uh Alphas to put the tissue into neutral or betas to put it into gear or fast betas high frequency Bas to put it into strong gear or Theta to take the brakes off and release it and then you can look at things like the speed of the brain so if your alpha waves are running faster then literally your speed of processing is faster you're thinking faster you're retrieving information handing off information inside your mind quickly and so the these are the arousal phenomena um essentially the the regulatory arousal not sort of sexual arousal but you can think of arousal in a context of how awake you are like a like a thermostat or tone that's where that comes from but it really is sort of the SMR landscape sensory motor Rhythm which is one of the core frequencies we use in classic neura feedback I would say that at least half the field is still centered around this one frequency and it's involv the regulatory rousel management so an inhibitory tone on aspects of action sleep Etc so when someone does a rousel model neuro feedback they probably do a qeg and they probably do multiple band training and try to tune different bits of tissue into directions you want under the idea that we have these you know patterns we can go after and resources we can help and then you'll often see that model that that arousa model heavily adopted by people that work in the autism landscape because there's a developmental kind of perspective to it as well because you can kind of take one kid who's got a lot of Developmental difficulties and based on a bunch of factors in their EEG you're kind of age uh pegging them little differently than they are and saying ah okay this seven-year-old without language and these tissues this tissue is working differently try to create a a developmental change in that tissue through the same model so that's the long-winded answer but thank you for asking and and we owe that to the aers aers sort of yeah I mean to some extent we do the SMR stuff of course comes from Dr Barry Sturman who first would of demonstrated that these this this reinforcement of SMR sensory motori themm creates a inhibitory response that allows brains to resist seizure and from there we discover that SMR is also heavily involved with sleep it's called a sleep spindle if you're asleep it keeps you asleep and the fact that SMR is a sleep phenomena and a anti-seizure phenomena and a sitting still phenomena sort of created that core arousal the aers were heavily involved from a business perspective and an evangelism perspective and a deep care for lots and lots of people an education perspective in the original in some ways commercialization as neuro feedback started to take hold and I think the 80s uh really is when they they got into it I forget exactly when it was before my time and maybe 90s and uh they they used a a three band SMR focused arousal model and the software had a different name at the time now it's called eager it's the same software that a lot of the field uses now it was called I think neurocybernetics before when I first got involved in the field to give you guys some sense of how long ago it was you need two computers to run that software and they were connected by a parallel cable and one ran the game and one ran the signals and you kind of had to have them talk to each other um but the offers then changed you know Sueur was finding that working lower and lower in frequencies was working for certain populations of people especially those with strong developmental overarousal the over aroused nervous system is that autistic kid who's screaming and stemming and can't self-regulate it's an over uh disregulated over aroused nervous system so Sue was finding that as she went to lower and lower frequency she was getting some good response across that complaint and also across I think a trauma landscape a trauma informed therapists were starting to get good results following you chasing subjective experience you know because Sue had this process she was she was never really uh super focused on qegs and went do an awful lot lot of again this is very true of the arousal model therapist and provider of of which I sort of am one as well I I was trained in this landscape and still mostly work in this landscape but but you you you tune so one person's SMR frequency is not the next person and and how you adjust it is in response to what they notice when training so if you train someone's I don't know you train someone's C4 smmr at 11 and a half Hertz and they don't sleep very well you might go up to 12 Hertz and get a better sleep effect and so you learn to watch effects these arousal effects regulatory effects of sleep stress and attention in response to these features so Sue was was doing this she was adjusting and and and tweaking and moving down and was Finding really good effects with certain populations and to the point where she was getting below one Hertz and the amplifiers they were using were having trouble resolving the signals well so Sig fre ofur did some engineering and helped you know did a couple iterations in the the called neuro amp I think um ended up having a a lower and lower floor for coupling until the point where they can get down to like 0.001 Hertz you know very long slow waves and so at that point uh the opers moved much more into that direction they created some new software there's a part of the story I skipped where um the software they had been using they sold to a different neuro feedback company that was trying to sort of take uh a big play you know try to make it large and that company ended up folding a couple years later and so the soft reverted back to Howard light Stone's uh um ownership and then he relaunched it under the name eager uh from the company eager and so that's what we have now still as a as a nice backbone for the field but the offers released um Signet with a c you know baby swan and uh Signet is infr low or infr slow I always get it confused which is which Mark Jones Mark Smith uh has infra infr slow as well and there's a few different people working in that landscape and it's not the same thing but it's also there's also regulatory components in that frequency range in the slow cortical potential sort of DC shift of the brain as well so there's several people working below one Hertz um you can look at Neil's beer bomber's work for slow cortical potential especially those are also arousal model and you'll find a lot of trauma-informed therapists do these in for low and for slow because they do uh or they can steal the nervous system pretty quickly so what allows the offers to train that low is of the software or is the hardware different because doing ilf with the quiz doesn't feel as good as with narrow amp 2 yeah the qiz um does resolve frequencies pretty low it is a DC coupled amp if you use it in DC coupled mode so there's a couple tricks to get a an ilf or an ISL effect that's really solid with a Q with a pocket aerobics amp and I believe this is true of the four Channel uiz I think the four channel uiz is the same amp just repackaged for anyone who's looking for a q is you'll find I think they're mostly discontinued and they they've replaced it with a very small like a small form factor here's here's the queue is for folks that are wondering it looks like this and there's one that's half this size with most of the same ports the same four Channel setup called the uiz and I believe the the same innards where I forget what the lower end is it's pretty low and the big complaint with or the big problem with doing infr low or infr slow with this Hardware is generally the software side you know a lot of people are using um bioexplorer for instance for doing the The Whiz amplifiers and the bio Explorer software stack only gets down to I think. three 0.3 Hertz before you just can't resolve it through the software tools that are built into the the filtering essentially so you have to a little bit roll your own and a lot of people will take bio era which is kind of like a non-user friendly version of Bio Explorer um it's more powerful but less configured out of the box and people will configure their own software essentially using bio era and that's what Signet was initially it was a version of Bio ERA with another skin on top of it heavily configured running on neuro amp so there's nothing really in the way of using the Q is that way but You' have to kind of configure your own stack because bio Explorer is not going to work and you know I'm not sure how your filtering is below that half a Hertz range where a lot of software will just not think about frequencies you also must use centered Center Silver silver chloride electrodes on the amp you have to use a DC coupled electrode if you use a flat silver electrode then you can never send then then the there's something called a i a electrical potential volt potential and what what's happening is you're getting fluctuating voltages and at the scalp and electrod is measuring that fluctuating electrode uh the uhal difference if it's a pasted electrode and a silver 10 electrode or gold electrode then the Electrical uh fluctuating signal is always on the outside of the scalp it can't go in this there's too much resistance with that kind of electrode but if you have a centered silver silver chloride and you have a driven right leg you know ground which The Whiz is due then you can get down into the DC coupled level essentially to zero or like the neuro amp to you know 0.001 but you're going to have to have software that's good at filtering at that level that's stabilizing signals that level um you know and and it's it may not be quite as focused on that task as the neuro amp uh uh amps are and so you might be better off just getting a signant system getting trained to use it by uh the auth by by Sig Freed's team if you uh want to use that system in that style or you know um eager just added a module I haven't tried it yet I haven't trained in it but eager has a sub Herz module now built into it that they're they've launched so you know they were they were late to the party for the infow infr slow but you can get below one Hertz Now using eager software if you're an eager user and you've you've been missing you wondering about the uh the below one Herz stuff you can do it now so G gamification we've talked about this in the past I don't think you're a fan you don't like zuk cor or you have other things I use zukor no no I I like zukor yeah I sell zukor I like it um people like it what I don't think is um what I don't like is an emphasis on the reward feedback the style I I I don't think it matters that much and so I think games matter a lot when someone's really bored um or won't sit still or you know has some other reason where they want a game like I I have adults that are like I can't look at a maze of pictures or you know Pac-Man what else you got I'm like oh I have the software package you know now zukor is somewhat expensive um so it you know it costs several hundred dollars to buy a buy a license of it so you know it has to be a good reason for it not just like I'm a little bit bored at the four or five or six games that I have that work fine I mean when you you you feel neur feedback when that you're excited to train you get effects you know you could ignore it you get effects so sitting there and staring at it and trying to make things happen and that expectation as aspect of reward learning is not not necessary for this flavor of instrumental conditioning because it's involuntary so you don't need the expectation doesn't really do much as far as I can tell and therefore the games can be kind of discreet and boring and maybe the information gets in a bit more cleanly I have observed with some pieces of reward software or reward environments where if you load things up too heavily you seem to get reduced impacts from the same protocols and I I really do think that it's true that if you use video Fe feedb back you can risk blunting reward or the impact of neur feedback very easily and I've seen it happen again and again and again I don't know why um exactly but my big theory is that you know neuro feedback is bound by requires the resources of implicit learning you gotta kind of your brain extracts rules from patterns it's implicit learning social information loads up implicit learning tissues with processing and floods them and you can't do back ground math implicit learning so you can't extract the information the feedback event see here's the thing I know it's happening in your feedback I know what the brain's doing after a beep and it's it's really responding to the specific Applause to the specific change it just made and if you have lots of really really interesting social things I'm afraid that that effect might be abolished I haven't tested this but there is research showing that social loading will abolish implicit uh uh learning so I think it's a little bit risky to use videos unless that's the only way the kid will sit still and then you're getting some training in versus some movies no I don't like movies I mean I offer them every so often I have to you know when when that's the only way that you you have a kid who has some significant autism who will watch that one movie 75 times in a row but won't sit still for anything else guess what we're doing you know we're watching that movie 75 times in a row because that's going to get the training in and if it's only 70% as effective it's still 100% more than not getting the training in so did you ever use that as a selling point I've see a lot of clinics including my own you watch movies and train your brain yeah yeah I've never used that as a selling point no not once I my my selling point starts before neur feedback it's a more like oh you want to understand your brain yeah I'm gonna teach you understand your brain I'm G make you your own expert in You by looking at data and telling you things and showing you how it paints out you're going to say oh my gosh yes that's me you know when you're doing brain mapping properly it's not mysterious you're like oh my God yeah I had no idea you could see that so oh got an order for the kitchen let's see he what the heck is beta reset what is it used for what does it do beta reset was invented discovered much with the right Languages by jacyn Ginsburg and I'm forgetting her business partner's name and they're I think they're Arizona initially I was trained by them to do it at UCLA years ago like 2006 or seven um it's a two channel bilateral low on the head protocol I think it's at like I think it's a P3 and P4 and io1 and io2 meaning you're on either side of the Indian at the same level as the inan but you're also in line with like the the the P3 P4 you know front to back so it's like it's like a it's like a bracket like this in the back of the essentially and if you have a copy of eager and you can open it up and there actually is a beta uh pre uh beta reset custom session plan you can load that's actually included by default and it runs through I'm going to say it's like 15 or 16 or 17 different little short segments it starts off at a three minute segment and it and then it goes into 90minut segments and and it rotates through and what it's doing is it's sweeping down and then back up through the beta and it goes up pretty high and often the theory is and people do notice this there's like a sense of like release or like a sense of like ah like a physical sense almost that happens during it so people use it for a bunch of stuff um I use it here and there I have a few clients for whom we've tried it and they enjoy it um the people that developed it will say that it's useful for body based things they've used it for um in physical injuries in the body they used it for cible paly with limb wasting they've used it for parkinsonian phenomena like deep brain phenomena um and there does seem to be some other effect it's not the same as band training it creates a different impact and it's probably doing something that's you know really around the disrupting hyper coherence kind of landscape at the very least you know the reset aspects while it sweeps through and sweeps back up that's going to really disrupt hyper coherence when you have that particular phenomena and since the back of the brain is the outside world including the body it makes some sense we get body based phenomena but it's not a well understood uh protocol it's more like a recipe book that you know like many in things in the field that we kind of get some effects from and they're kind of predictable but it's a little bit of a protocol that's hard to um explain what's happening in that one yeah up order up how important is it during trading to engage with the feedback like in children with autism is it hard to have them engaged is just audio feedback enough while they are playing a game yeah you don't need any engagement whatsoever literally people can be in a coma literally in a coma and it still trains the brain Margaret airs um for years at the you know her last part of her career she would go to coma bedsides and train brains to be more resilient reduce seizures that kind of thing so no it's an involuntary process mostly and it was discovered on cats cats are really bad instruction followers right so it's not this voluntary expectation it's more like a a passive reward but the brain's getting Applause contingent on what it just did so it has to do a particular thing and then that gets applauded sort of outlined or reinforced so um as far as I know it doesn't require any real Focus auditory work training works almost as well as multimodal training is my take so I do train mostly with Visual and auditory and a big effect is the auditory component like I'd rather train just auditory that works okay I wouldn't want to train just Visual and if somebody for instance was you know thoroughly deaf I would use a tactile strip to get a tactile response to make up for the lack of auditory if we could so I think a multimodal response is useful but the auditory is uh powerful and and and I have trained autistic kids that were deaf and blind and they would have a rumbly little teddy bear in their lap that would just vibr purely that one stimulus and that also works for things like seizure reduction and some other sleep issues so um I don't think it's necessary to focus uh or try and I have a ton of you know historically over the the decades I've had plenty of teenagers in my office who like in spite of what they were told would sit there on their phone and ignore it and like you know not pay attention and it works fine for them too some of them are are CEOs that are 50 you know some of that type who don't don't want do what I tell them just do whatever they want to do while they're doing training still works so you know on email on the phone the whole time and that should that should like get in the way it doesn't really I mean maybe they get a little extra effect uh by sitting in quietly and watching it but it's not very much as far as I can tell so so what is like the step by step when when somebody comes in I mean I don't know want to start with autism but you know that you have a parent and a child and I know you're doing a lot of your stuff remotely but you're looking to get some type of Buy in or commitment or uh to so they'll continue the process because it's many many many sessions how do you start them out to show them that hey you know what you do have uh control over this is it the the the temperature or what how what do you do I I don't do any peripheral bof feedback or or any sort of like voluntary bio feedback we really just say hey here's your brain check it out here's a bunch of ways your brain is built isn't it cool you didn't know you could see all these things about your brain you know are any of these things things you want to you know exercise or change so you know if working with like a non-verbal autistic kid or something it's more about their parents you know where are the goals where's the suffering you know what would success look like and at the parent level people often just want to know that we've worked with with complaints like their their kit which is you know pretty common when you have a lot of kid suffering you have several flavors if you will of kids suffering that shows up so those of us that work with kids see a lot of the same things again and again but with like a teenager who once who's going to show up and maybe have buyin or maybe not you can do all that before they get to the commitment to ner feedback you can literally do it over their brain map show an ADHD kid or an anxious kid or someone with some social loading or some sensory IR ability you show them their brain and you validate their experience and say oh look at that your brain's doing this that's interesting and if it's annoying that's frustrating oh sorry dealing with that but look at that you can see it isn't it cool you can see your rumination you can see your tick right there is that cool you can see your ADHD huh and just right then without even doing neur a feedback we end up with this transformation you know on podcast I've mentioned once or twice that I've gotten a story here I got an email yesterday from a dad who said hey just had to tell you since you had that Q meeting with my daughter my she's 18 19y old daughter a couple weeks ago it's like everything changed she started studying she started organizing her time she just she started focusing on College and how to prepare for it like you know just seeing her brain gave her the permission to start working with her brain and you know your suffering is not instantly different but it's instantly harder to be overwhelmed by anxiety ADHD sleep issues whatever if you see it as physiology as a bit mechanical gives you another mode to start you know being gentle with yourself and working through resource instead of having a diagnosis happen to you so that piece of it actually is really powerful with teens who feel disempowered overwhelmed frustrated so that's my my my sneaky trick is I just provide massive agency through education and people want to do stuff even teenagers generally radar is trying to turn neuro feedback into the next online Health Trend you're preaching the choir here I think take a number that's right that's right what what do you think is gonna have to happen is a couple more Netflix series need to come out or a couple more sports figures need to come out what do you think I think the complexity of the process will have to go will have to be reduced there's there's a lot of people in the fringes of neuro feedback that are geeking out around um computational optimization in neur feedback doing AI work doing database uh driven stuff and it's super early stages but I'm seeing people create rather compelling analysis rather compelling uh training plans using computational methods and not just individuals with deep skill doing it so I think that's going to accelerate and I think that within two or three years we're going to have the ability I may have talked about talked to you about this before Pete but I think we'll have the ability to have a mathematical model of our brain of our Wellness of our sleep of our addiction of our mood in in you know in silica and we'll be able to do things where we can sort of model the impacts of interventions on ourselves before the intervention happens and massively you know narrow down the types of things that will be useful for our brain and our body and then provide tailored neur feedback without people like me in the mix so I think I think my I think my uh my role in this world as educator is going to go away once the once we understand the brain better and can get the software you know learning a little bit about that stuff I think we need that Tony guy from Apple to come over and design this product because I think we've had Decades of narrow feedback being marketed and created by engineers and Engineers love to use it but anybody else they're like what the hell is this yeah definitely the current tools are still cutting edge you know late 1980s or something like they're still not really wonderful most of them I mean eager is not the sexiest software but I stick with it because it gets out of my way it's kind of a Swiss army knife it has all the tools I use um but there's no you I had somebody complaining a couple days ago like oh I can't you know working through a bug or two or some some setup issues with like 60 HZ noise because the wire wasn't on properly he scaled the software wrong or something and he's like getting frustrated and yeah it's a little frustrating but there's not a lot of pressure in the field to create to like iterate the same way you would if we were Apple you know Apple's would would throw a lot of money at this and have really nice software very quickly because the software is not that that complicated know somebody with $10 million could like own the software landscape if they decided to like that they wanted to own that you know 15,000 install based there's not that much it's a it's we're a little niche little tiny niche of providers and a bunch of prosumers doing stuff that's kind of it so that's the problem radar we have to we have to go viral on substack as you say or Twitter yeah yeah how do we do that how do you how do we do that we you know what viral is up yeah I don't know we have to get me in front of like big big names when I was on Joe Rogan yeah I got calls for the next two months from providers saying hey thanks what thanks um I just got a big rush of business so and and I did too you know it was a pretty big large impact uh you know for the next couple of months we got dozens of people coming in we gotta get you back on with Joe how'd you get in there I forgot how we get in initially I just asking I think but you know uh I'm not sure that he's my biggest fan so you know I don't think I'll back why what happened I don't know it's one of those conversations where the end of it wasn't you know as as good as the beginning uh oh come on I get those every day right I know uh but no that's I mean we just need to evangelize people just need to know that you can look at your brain understand yourself make changes you know people don't know that you can do this stuff they don't know that you can look at your executive function or your stress response or your speeder processing and you can see it so qeg I think honestly is too complicated too messy too expensive when that gets you know that piece of it still has room to come down the the the neuro feedback software as an industry because it's nichy it's kind of locked up in its current form it will go through evolutionary changes we'll end up with very inexpensive very high quality software five years 10 years and that's coming because software development is changing rapidly so somebody who's not in the near feedback space will just swoop in and do it if the current providers don't get better uxui run them you know run on on smaller Hardware Etc so that's software is coming but qeg rapid data acquisition rapid processing cleaning that's actually kind of a hard process looking at data and when that gets less expensive when you can send somebody a couple hundred dollar device passively read it good to go and it's not several thousand dollars of of outlay several hours of time you know when that piece of it drops I think that the current landscape will explode uh doesn't it but look what the software companies do look at Salesforce look AI they they go to the golf courses and maybe people won't use neuro feedback to make their lives better but they will spend a gazillion to drive their score down a point yeah why can't we get somebody in into the golf course is or nobody's got money I mean there are people who focus on golf as a primary business in neur feedback you know there's several uh athletic neur feedback focused people we have one in at Peak brain over in our UK office it's doing a lot of athletic focused work but um there's also individual products I mean John Cowen the play attention guy um he developed a version like like a like an oem repackaged version of The Whiz and well before the whiz the the the peanut or the pendant or something and it was built into a golf visor and it had audio feedback for like dropping into the zone so when EMG dropped and SMR climbed up so you would like square up on the you know on the te and then like deeply relax and you get an audio pitch climbing U so for years he had a whole business line which was focused on following go and there was a laptop nearby too on Bluetooth so it was like you know follow golfers around with a laptop you'd hear some audio happening so um but if you had a simulator in your basement why wouldn't you have like it's true right yeah and there's and there's places like I I live here in La there's places in La where it's like a little room that's a little fake you know Golf Course for for teaching people to to swing so you could put it right in there absolutely yeah I I should partner with golf education or golf uh what's the term I don't I'm not a golf I don't know I last place you'll ever find me doctor is on the golf course no we should on one of podcast can see you're rocking a pair of those those plaid oh yeah stop at that's right that's right no but I think that's a great idea we we should get somebody on one of the shows to you know talk about golf because we what's it Brian uh I'm gonna butcher his last name bo bo Chomp uh he he's a big proponent um we had uh there been individual sports uh athletes that have been you know talking about Fe back with the Netflix thing I have a hockey player named Connor Carrick who got kicked out he he got kind of injured out of uh the league he was in and during did a bunch of rehab in the minors and work with us and he would like train his brain on the bus between you know scrimmages and practices and things and worked himself back into top form and he got signed by the O Oilers last year and he's been kick and butt so you know individual athletes are often like n feedback everyone should do it but sometimes athletes until they become major major Stars aren't don't have you know have the most impact necessarily as a as a voice so they cousin I mean he he he busted out there he was on Netflix and then you know I forgot the name of the company that was sponsoring you know the the headband but wasn't even neuro feedback was an HRV device you know was HRV or was it HG I forgot what it was I think it was HRV just breathing that's what they were feedback yeah yeah the bio feedback uh end of it but kurk cousins would use neuro feedback but the company that was promoting them you know that that was HRV I mean however you started to get it going I just can't believe you know is there is there a higher uh Revenue per year in copper bracelets or neuro feedback probably copper bracelets in terms of the margin yeah yeah we're in the wrong business doc I know so so here's the thing though you know like right now I use a few thousand dollars of gear for each person I use you know a few thousand of assessment equipment um we do have some devices starting to come out there that do things like real time qgs and they do HRV and PBM you know things like Divergence neuros headset for instance um and there's been some plays like the sense AI which is trying to do something where it's sort of semic canned um but I don't think these devices yet have solved these problems in the case of like Divergence neuro and some of the other devices out there the iyn wave some of these devices are really really expensive still so while they're starting to S to solve a problem there's still so our our field is still so early in some ways that there's not Market pressure economics of scale to take a device down to prosumer cost so to I think we've been trapped and this is what I've been seeing for 20 years we do have some really interesting devices that pop up get used by some people and then drop away and like there's all kinds of really interesting devices out there that no one knows about you know mag stem devices and interesting you know the the roshi 2 I'm thinking of we had a MAG stem and a light stem and an EEG all doing feedback at the same time you know really interesting stuff but the field has no Market pressure so you know what's what's uh Jay say you know how do you start how do you make a a small fortune in neuro feedback start with a large one start with a large one exactly so it's so nichy that you know we're gonna need someone like uh you know a top level influencer to make neur feedback sexy against the current model and then have huge adoption or we're going to have to avid evolve into the place where the software the hardware and the supervision become a lot more accessible and it's not a couple thousand dollar a month to do it which is what you know I charge a couple thousand dollars to to to get started basically it's a really big barrier for many people so it's it's kind of a hard uh kind of a hard thing to start into the you know and and like you said at the beginning of the talk today the the show today you have to get some buyin because people don't feel Nerf feedback you don't get change immediately you don't feel it like a drug sometimes you can feel quickly so you're two weeks in two and a half weeks in three weeks in and there starts to be these little effects and so I just set expectations that way let people know that you know takes a minute it's like working out you you know builds up so it's it's hard though you're you're you're definitely putting your thumb on some things we haven't yet solved in the field and and I gotta tell you it's a it's a little frustrating because I get asked by clients routinely at least once a month someone asked me you know the the the context is okay oh my God I feel amazing what is going on I feel really really good oh my gosh look at my attention test scores oh my gosh look at my brain Maps my wife likes it I feel really good what is going on and why isn't everyone else doing this why doesn't everyone know about this that question everyone not everyone but I get that question a lot and the qu the answer is well it's a little complicated to do it's a little expensive it's a process takes a little bit of skill bit of knowledge and all that can conspire to keep the barriers to entry a little high but I think I think many of those things the costs and the complexity I think will come down uh in technology we you see one of the Kardashians to uh go around with a im medicin helmet and right whatever something yeah I mean I did a I did a EEG on Tyler Henry the Hollywood medium on Dr Drew's show or actually vice versa Dr Drew and I went into Tyler show and we did a a play byplay while Tyler was doing his uh his psychic reading thing on steveo and and we got a whole bunch of people coming in over the next you know year or so that were like psychic Affiliated you know people that were interested in in that end of the pool and uh you know had sort of like were followers of Tyler or other psychics were like oh that was cool can you look at M now and that's how I got to look at several people doing you know unusual experience work in their EG because they saw that one that one show so I mean I find that people don't know that it exists and if you explain relatively simply what it is they always understand and then they always want it so it's almost like you got to Market it as something else besides mental health and then the side effect is it helps your mental health you know yeah well it's really close to to personal training and how it's done it's it's agency you take through assessments and that's it's iterative change you make with feedback subjectively generally not every form is like this some forms are self-driven or canned or one- size fits all and in some clinical models the provider does all the decision- making the process itself is one that's iterative Progressive relies on your feedback goes after your goals and it's not as tightly tied to the diagnostic landscape or the fixing diagnosis perspective than you might think and so I just you know again teach people their brains and that gives them all the agency I find that I have a pretty good you know follow through and compliance with things like training and Reporting and talking about goals and it it goes up when you teach people about the process and about themselves when you tell them why things are happening their behavior changes faster than just telling them something to do so you know that's how I bring in the expectation piece that radar mentioned is not in the game it's in like ownership of the whole process setting your goals what' you notice afterwards and let's rinse and repeat Anthony I gave you the link buddy hi Anthony from other side of the screen he outside looking in I think the variance in phenotypes is why none of the all-in-one devices take off yeah the that you can have the same phenotype with different meaning you can have different versions of different of the same phenotype absolutely and it is hard to spot I think it's very very SC very um astute Anthony I I saw somebody post recently on uh one of the Reddit you know we every day somebody posts a hey explain my brain map kind of post on Reddit and in spite of me saying hey know please post some actual data and enough pages and some raw get these like you know short little onepage ones every so often and one popped up it was full of noise and it was obvious but like even through that noise I'm like okay this person's got sleep maintenance issues oh some brain fog so like once you've looked at thousands of brain Maps even if it's noisy data you get kind of good at it as long as you have the raw data to like fall back on to know if you're actually looking at real stuff so so the phenotypes are subtle but I have to believe Anthony that we're going to get to the place where machine learning algorithms can get them as well or better than than people like Jay or me who you know and and and I I'm I'm being lofty putting myself in Jay's category of looking at brains but like I've looked at a lot for you know a quarter Century so I'm gonna give myself some prop for younger right right for for a youngster I've looked at a lot of eegs and um I I don't think that's going to matter as much I mean I've got you know so many EGS and so much learning and review but like that could easily be replicated in in a semi-intelligent landscape not even modern AI just machine learning doing pattern matching you know and not even on the on the maps necessarily but on the wavelets the waveforms on the distributions you know more modern version of qag would decompose the raw into the into the values essentially right in and interpret them as that happened not get to pictures and then work backwards which is often you you brought up machine learning have you gotten into deep seek at all I not I've not tried deep seek yet I I've gotten into it I've uh I kind of have my uh my group I have my GPT I have my deep seek and I have my grock and okay kind of what I do is I don't know if you see it on the social media side to get some people an ey um I take a Hot Topic and I have I I say chat PT uh you got this side you got that side and let one of them adjudicate you know so you can see the process the thing I like about deep seek is it goes down the critical thinking process of what they're considering as it comes out I think you can turn that on now and it's it's it's a thing you that GPT was hiding because they were afraid that if they exposed that that other models would learn off of them and create versions and it turns out that it doesn't matter if you hide it you can still distill the model which is probably what deep seek did it's why it was only a$6 million dollars because they probably distilled the logic by talking to chat GPT there's some evidence that they there was an article about how there was this massive API access from China into chat GPT doing huge amounts of question and answer to look for the like like the logic chain probably so D I'm getting way ahead of my skis here okay so you you gotta you know catch me with butterfly nut but I you can write script this is stuff you probably already know but I'm just looking at for the first time you can write your script that says I want to do this I want to uh whatever medial task that you're doing somehow you can automate it whatever it is but then you have two other um advisors that you can go to to figure out if it's right or not or to see where the law is in the script reinforcement learning the the big difference is the reinforcement learning is coming in now so we had you know predictive algorithms up until now all of the large language models are just predictive models you put in input you get an output but they were very very bad at tuning the output in real time at test time or at at at uh inference time and a lot of what's being done right now is to put working memory to put a reflective period before answering and then to have reinforcement learning built in to tune the model's ability to get the answer and that is something that that deep seek seems to have done better than chat GPT was building the the uh reflection at answer time process uh and that's probably going to keep going we're probably going to see all of this now you know fairly accelerating so um well and way cheaper way cheaper no I mean there's there there's there sort of a paradox right it went cheaper but what that means is we'll just use the same technology now we'll use the same money we'll just do a lot more with it right like it's not we're not going to you know if it's if it's 5% of the cost and GPT open AI sees this and goes oh okay this style of doing it is 5% of the cost now they're going to implement the same thing because deep seeks uh open source and the same shortcuts and then they're going to do 20 times as much computation for the same money they're already spending the same resources and power and cooling they're already using so it's just going to allow a fast growth is my take on that and here's the thing deep seek is kind of interesting I if you saw the article yesterday day before somebody sat down with deep seek R1 and challenged it to find ways to improve its speed and it went out and did a bunch of things and came back and it doubled its speed literally doubled its speed of processing without any programming just a a AI developer saying okay I think you should be able to figure out a way to run twice as fast or run faster so hey radar I'm not a neuro feedback practitioner I was I started with a large fortune and got to a small fortune and I said here I am on you know doing the podcast my sister was a neuroschistosomiasis and that funny I know a lot of talk therapists that move into neur feedback because they get tired of sitting in a room having one person come in and come out every day and only having limited contact and going through 30 hours of that every week instead they want to have a nice little busy office a couple of stations and people coming and going and you know it can be really boring and really kind of isolating to be a talk therapist in private practice and so I know a few people that have expanded just because it means have a nice you know multi activity office and so it gets things really you know more more culturally You Know Rich there so so doc I don't want to get too much in the weeds it's only you and I so we can really get into it stop us okay to stop us but isn't this machine learning isn't it all about an image and looking where text is or a question is asked it all comes down to being like a picture right yeah pattern matching P PCH pattern identification you can do all kinds of different machine learning when I was I worked at mlan hospital for a couple like a year and I was working for Dr Mike Henry who ran the ECT uh Center and did a lot of um uh other work with depression and he was running some studies looking what happened when somebody withdrew from their medication because people come in and say look I don't be on my anti-depressants anymore and they would withdraw and we would scan them in an MRI several times over a few months and I built a machine learning algorithm in the software SPM to fall the edge of the structures and Trace itself and stay on the edge so I didn't have to sit there and like well for the first several thousand hours I sat there 100 hours and like Trace by hand the cod8 the pment ETC but after a while I learned to do a bit of scripting and created a machine learning algorithm to like you know trace the edges of structures based on the light gray values that it was seeing in the picture so yeah you could sort of give it feedback and get better by you know telling it what it was doing right or wrong and these days modern things if you can Define what the answer is what the right or wrong is then you can kind of give it infinite ability to learn and self-reflect and this is how things like deep seek we're able to go oh yeah I can be twice as fast boom now now the code's twice as fast so where I'm going with it on the image part of things I mean that's ra EEG right you should it you have to be able to read it with a machine yeah that's really wavelet analysis right the shape of the waveform uh within a certain amount of variants um and yeah there are some initial uh uh visual scan EEG um systems u Lisa Taran was was reading in raw EEG into chat GPT she built a custom chat GPT which has her you know 20 article database of things she uses for analysis and the uh mat lab EEG lab manual of all the things it can do and then her pre-processed files that she does out of uh uh mat lab and takes those and runs them through chat GPT and it does a signal analysis it takes EDF files and mat. matat files so we're not talking PDFs anymore we're talking EEG files you know binary files takes those says okay there's some Alpha slowing here okay this area has some of that there on machine learning in this case and that's not just predictive outcomes now we're doing the ability to reference uh resources in other uh contexts so of essentially have a working knowledge base that is then brought back in I'm working on that myself now for a tool not to do the work but to summarize what I've done because I'm not convinced that like Anthony said those phenotypes are pretty subtle sometimes I'm not convinced that the the AI is be good at brain mapping reviews for a while but I am convinced they can take a 30 minute video of me identifying all those things and talking to somebody about about goals and brain mapping and what it means and to still that out into a much more regular explanation for somebody to then digest later so well put a bow on it I think that the three options that you have that you can use GPT deep seek and grock and then with uh Raspberry Pi these computers are getting so small I I have to think that you could at least maybe can't you just measure one spot to get the signal off of brain by by using that for you mean as a proxy for Q you could what you'll need is more of a wild type database the the qeg landscape to to make the predictions we make we have to control a lot of variability and a single electrode has too much variability to judge unless you have a huge amount of context so this is why you can do things like like two channel zscore training at a location in software you don't have to you have to you don't require a full head or Loretta analysis to do zcore training you can do it at a location using a zcore DL or something from brain Master let's say but to really do signal uh discriminants let's say to really see signals across other signals you need more than one electrode you really need multiple electrodes so things like coherence and phase for instance you're never going to get if you only have a couple of electrodes this sort of gets back to the idea there are some things on our field that use language that sounds an awful lot like qeg and it's not because they're using like three or four electrodes in the front of the head or they have sequential pairs they move to do one minute of recording or two minutes of recording at a time and those things also rapidly lose most of the of the utility of qeg because they're not getting the intra electrode simultaneous measurements of patterns that are that are existing so I think we're going to be constrained for a while of needing more than a couple of electrodes to do assessments but I do think that someone's going to come up with like a good dry system finally or a good you know zero contact system finally at some point so I expect we'll have a few technological breakthroughs as well that are not about neura feedback but it might be about like Material Science or electronics and then we'll end up with a Raspberry Pi or a phone app that can do neur feedback in a sophisticated way that's as good as the clinical tools but that's a much shorter jump to get to than having the tools that'll sort of assess without the qag being you know fully full gather I think I I guess what I'm hoping for is if you can get somebody hooked on just getting one point and showing you have control of uh CZ right um I forgot what the protocol is but that's the one that my sister would always go to you know somebody had go to CZ and put it on the top of the head and then you know see see what see what the outcome is if somebody gets started on that with a Raspberry Pi and you're talking about a couple hundred dollars out of the pocket I know that's less than I mean you want subscriptions you're gonna cause trouble yeah you're G you're gonna you're gonna cause harm you know someone you know two-thirds of people that's fine for or c for SMR two-thirds of people three qus that's fine for but like that last portion will have AB reactions and looking at a whole brain map will tell you what to work around what not to do how to should start perhaps so it's a little I think it's a little responsible to train blind I really do think we need a full picture as much as we can get these days which you know it's 10 minutes to 15 minutes of a cap on your head put a cap on your head squir it full of gel sit still that's it like come on like you know people complain about messy hair but I I don't want to hear it get the brain map know thyself it's worth a messy hair and what I'm saying is with these cheap things can you just put the one spot and come up with your own quasi yeah I mean maybe maybe eventually because remember part of the thing about EEG is you have to have the same Montage to judge the same EEG so you're going to need to be able to map reduce density electrodes into a larger density space which is going to be hard to do one won't do it a few might do it um there's a company out there making something and I'm not sure I buy the technological not sure if they pulled it off or if it's vaporware but it's an inear device that has several electrodes around the edge of the ear um like like a hearing agid style you know or or audio audio file earbud kind of musician you know active earbud kind of thing you put in it has a few little contact points around the edge and supposedly they are doing a Loretta like process to visualize where EEG is elsewhere now I don't know if it's working yet but the you know and I would normally think you know based on what I know about Loretta and sourc ization EG and what happens as you get out on the ears themselves I would normally be like probably not except that several years ago when Apple released the uh the the uh whatever their their their headphone is called airpod the airpod airpods yeah yeah when they released the second version of the pro initially in the in the um uh patent there was all this language around bio sensors embedded in the device picking up EEG so I think that it may be possible technologically but I think there's a a use case implementation like you know you can't just have an electric car you got to have all this infrastructure and you can't just have an EEG you got to have referential databases and analyses and qegs transformed and it's a whole landscape that you have to have to match the way you're measuring and so I think that piece of it might be solved first some might come up with like a really low density you know you pair glasses you put on it measures all down the side of the the head and behind the ears and somehow ends up with like a nice spatial visualization of stuff that's really kind of far away through the combination of lag and instantaneous EEG which is what Loretta is I think that might be possible but I don't think we're there yet is my my big takeaway so soon you know maybe next year all right last question and we'll shut the kitchen down doc speaking of Lisa I think she trains people with Parkinson's isn't 0102 used for people with PD what's the benefit of that I don't she trains uh Parkinson's yeah Lisa is like the the global Parkinson's expert Lisa Taran th yrn she's like if you if you're interested in Parkinson's and neuro feedback you should talk to Lisa um she's been funded by the Michael J fox Foundation she's done lots of interesting pilot studies and research um I don't know about 0102 radar um I think 0102 could be used for parkinsonian complaints if you're using 01us 02 except for the C Cel cerebell nah you're really in the visual tissue you might be getting some cerebella signal but not much because it most of the EEG and the cerebellum points the wrong direction and you can't measure it from outside the head so um but what you might have is beta hyper coherent across the back of the head in various degenerative diseases headaches you know various neuro things you'll have beta hyper coherent across the back of the head and that would break it up if you reward low beta or something and that with a difference Montage like that you would definitely you know break up some of the beta I find that I get my most strong impacts in parkinsonian complaints using SMR categories and I've seen several times classic SMR training usually mix of C4 and CZ um making medication work dramatically better where somebody with PD doesn't have to wake up in the middle of the night to take more meds and go back to sleep because they were tremoring where the meds now work all night long or someone who's young with a par andonian phenomena it seems to like slow down and stop progressing and I've seen the most impact on not on Tremor but on the rigidity the stiffness the woodenness and in face the micrographia these are all the negative symptoms of uh PD the the reduction of tone those seem to respond a little bit better um and then the dopaminergics the medication gets more sensitized like I said the same way it would if you were taking Aderall and you did neuro feedback you would get a stronger ader Aderall effect so I don't know if Lea is using 102 um I find that I don't usually need 0102 and stick to SMR style training now again dialed off their brain map and you can usually see issues in Theta beta tones in the C sites in the in the central strip on a qeg when there's proc andonian complaints so you know look at the qeg that's my short answer okay good Haida radar thanks for showing up again everybody else Tony for popping in Jay I know you're floating around Somewhere Out There russing up for tomorrow morning Andrew Hill for the people out there that want to get uh an atome EEG what do they do yeah so of course you can come see us in our physical offices which are New York City La Costa Mesa St Louis London Stockholm Nova Scotia or you can do remote work and in the US you can borrow equipment remotely coaches walk you through it so come check us out at Peak brain institute.com you can see all the locations when you're there or check me out on socials at Peak brain LA or YouTube at do Hill D RH L and uh come hang out on my live stream on Monday nights and ask questions yeah if you don't like to cook in here you can go Monday well you know I don't I don't just go to One restaurant or one takeout place I don't know about you Dr Hill man thank you so much for showing up awesome time nice to see you see everyone thanks for your your contributions tonight hey and uh keep that garden hose going out there it's coming you know coming from the sky I put the order in it's working it's all right we'll take it all right doc byebye right see you