Michael Breakspear “A theory of cognitive function based on corticohippocampal waves”

Transcript

• 00:07And I'd like to make time for organizing
• 00:10and and the smart ticker and everybody else
• 00:14who's organized this functional meeting.
• 00:16And yeah look I don't normally given you
• 00:19talk because there's just so much work
• 00:21but I thought the Todd and you guys,
• 00:24I'm going to write it for from scratch.
• 00:26It's a long plane ride and there's a
• 00:29long lonely jet lag nights being awake.
• 00:32Alright, this is you.
• 00:33I just finished this book at 3:00
• 00:35o'clock this morning. Sorry,
• 00:36it's it's probably going to be terrible.
• 00:41Turning out good. So
• 00:44we're going to talk about an experiment.
• 00:48Since we do do those sometimes still,
• 00:52then I'm going to talk about some
• 00:54pretty complicated analysis.
• 00:55That experiment which could
• 00:57easily occupy the 30 minutes.
• 00:59I have to run through all the details
• 01:02because it's really just using a couple
• 01:05of snapshots from the results to.
• 01:10Motivate my theory of the brain how it works,
• 01:14which is hopefully not that complicated.
• 01:16There's an Australian so
• 01:18can't be too complicated.
• 01:20So I'm gonna start with this experiment.
• 01:22This is work I did quite a long time ago.
• 01:25In fact, I worked with semen on this paper.
• 01:28We have a collaboration,
• 01:29but this was with Christine
• 01:31Gower and New Don Ren.
• 01:33So our interaction with the world
• 01:35continually informed by memories,
• 01:37right, which you retrieve
• 01:39naturally without mental effort.
• 01:40Sorry, you're going to see me.
• 01:42If I walk out the door and I
• 01:44come back in few minutes later,
• 01:45you go on this final breaks,
• 01:47alright, he's giving a talk and
• 01:49we don't really think about this,
• 01:50it just happens.
• 01:51So there's some sort of breaks in
• 01:53the stream of consciousness and then
• 01:55there's some way of gluing it all
• 01:57back together a few minutes later.
• 01:59So imagine you're watching the news.
• 02:02You go with this advertisement
• 02:04and you go to the fridge,
• 02:05get across the border or you get put
• 02:08in a scanner. And then after the ads.
• 02:13Comes on the rest of what you were watching.
• 02:14So in this case, we wouldn't really
• 02:16have an ad, but there's a news
• 02:18reader in the in in The Newsroom,
• 02:20and then there's a break.
• 02:22And then we put people in the scanner.
• 02:24And then we watched two types of movies,
• 02:27either those that were the news item
• 02:30that really newsreader was reading
• 02:33out or something completely random.
• 02:35Another news clip, I should say.
• 02:37So his new thread is saying, you know,
• 02:40terrible storms in Brisbane today.
• 02:42Uh, that kind of ****** the
• 02:44reporter on the ground.
• 02:45And this is the continuing condition.
• 02:48Yeah. Thank you.
• 02:49I don't know who this guy is,
• 02:51but he's American comedian.
• 02:53Thanks, Ted. Yeah, very bad.
• 02:55Brian here.
• 02:56Or it goes to this picture of a Pussycat.
• 02:59Yeah. Firemen were called to,
• 03:01to to get this cat out of the tree.
• 03:04This is the sort of it could be.
• 03:07And and it's all counterbalanced.
• 03:08So sometimes there's clips
• 03:10of the continuing 1.
• 03:12These are results in this paper was
• 03:15from small number of healthy adults,
• 03:18but then we put this into a big study
• 03:22with 186 healthy adults and I'll also
• 03:24show some people with mild cognitive
• 03:27impairment and Alzheimer's disease.
• 03:29So these effects are really strong.
• 03:31These are statistical maps,
• 03:33right?
• 03:33So I'm gonna get with Tim and
• 03:36say it's not necessarily a lot of
• 03:38great activity that's changing,
• 03:40but if you put 186 people into the scanner,
• 03:43you're going to get even with
• 03:45weak changes in brain activity,
• 03:47you're going to get strong
• 03:49statistical affairs.
• 03:50So this is like family wise error
• 03:52corrected at the Vauxhall tenant,
• 03:54minus four or five or something,
• 03:57huge effects of watching the clip.
• 03:59Compared to the rest breaks in these
• 04:03visual product rattle executive
• 04:05areas and then if you do the
• 04:08reverse GLM we see the default mode
• 04:10and that's not that interesting.
• 04:12But what's quite interesting to me
• 04:14is that if you do the the reverse.
• 04:17So if you just take the film clips
• 04:19and I'm just showing you this to
• 04:21convince you that something quite
• 04:23interesting is happening here
• 04:24if you look at the naive Cliff,
• 04:26the Pussycat and the tree compared
• 04:28to the storm.
• 04:29Are you do you see this very
• 04:31localized effect?
• 04:32This is the only effect in the brain.
• 04:35Other statistical threshold.
• 04:36This is like the something like
• 04:39the parahippocampal place area.
• 04:42I call this looking for the sock network,
• 04:44you know, you know,
• 04:46you're looking at the film tip going.
• 04:48What's going on here?
• 04:49I mean, a lot of people have this
• 04:51response when they're hearing me talk,
• 04:52actually.
• 04:55I'm always looking for my sock right?
• 04:59And then I find it,
• 05:00and off to work I go 5 minutes later.
• 05:05Response so.
• 05:08Yep, sometimes with one shoe.
• 05:12And this is the reverse contrast
• 05:14and this really strong effect.
• 05:16The clips are the same.
• 05:19People are just flying in the scanner,
• 05:21but when there are these.
• 05:23Familiar semantic clues.
• 05:25It really switches this looking for
• 05:28the sock network into this big.
• 05:32I guess you could call it the engram network.
• 05:34It's like, oh, OK, this makes sense.
• 05:37I've got an autobiographical
• 05:39connection with this film clip.
• 05:42I'm going to. I'm going to understand
• 05:44the haircut of what's going on here.
• 05:47And if you do a, the hippocampus is in here.
• 05:51This is a hippocampus mask,
• 05:53just to show that there's
• 05:55effects across the hippocampus,
• 05:56particularly in the head of hippocampus.
• 05:59I said this is all interesting,
• 06:00but not that interesting.
• 06:02This is kind of what I showed.
• 06:05If we're looking at the cortex,
• 06:07these are boxes fly.
• 06:08We just gotta model up with the GLM with
• 06:11these two different direct task effects.
• 06:14But what we did instead at begin with
• 06:17is we put this into a API equation.
• 06:20So we said, well,
• 06:22they're boring, we'll get them,
• 06:24we won't pay much attention to them.
• 06:27We're going to have a look at the
• 06:29activity in the cortex that we can
• 06:31account for by activity in the hippocampus.
• 06:34This is the baseline.
• 06:37Functional connectivity between keep
• 06:39careful variance and cortical variance,
• 06:42and these are these two interesting terms.
• 06:45So how much of what's going on in the
• 06:48hippocampus during these two tasks
• 06:50accounts for what's going on in the cortex?
• 06:53This is Parker physical interaction
• 06:55or whatever you want to call it.
• 06:56So does everybody sort of get that?
• 06:59We take away the task picks, take away,
• 07:01functional connectivity and we're
• 07:03looking at these tasks dependent modulations.
• 07:08So these are quite huge,
• 07:10but not that huge in some instances.
• 07:12So this is the baseline coefficient
• 07:15in the hippocampus.
• 07:16I've showed the cortical,
• 07:18these are the eggs.
• 07:20And so that's the taste of fistic.
• 07:22And then make a threshold that we get a very,
• 07:24very, very, very strong pay value.
• 07:26We can look in the cortex.
• 07:28So few things with time here.
• 07:31So first of all,
• 07:32yes, Tim,
• 07:33baseline much stronger effect
• 07:35is much weaker task modulations.
• 07:38Let's see the top base baseline effects.
• 07:42But what's interesting here is
• 07:44every hippocampus has a functional
• 07:47connection with some part of the cortex,
• 07:50and most parts the cortex have
• 07:53various that can be explained by
• 07:55the variance in the hippocampus
• 07:57particular to that task effect.
• 07:59So this is a better way of looking at
• 08:01brain really this is a whole grain.
• 08:03Everything's kind of running away
• 08:05through like your harness talk,
• 08:06there's big spatial modes.
• 08:07If we get enough data,
• 08:09we can sort of start to say that,
• 08:11OK,
• 08:12but then we wanted to say what's
• 08:13the texture of this effect?
• 08:15So we all know this work by Daniel
• 08:18on gradients.
• 08:19This is a purely sort of cortical
• 08:22decomposition using a nonlinear
• 08:24dimension reduction technique based on
• 08:28functional connectivity in the cortex.
• 08:30More recently, people have
• 08:31looked at the hippocampus.
• 08:33You find these same
• 08:35gradients and hippocampus,
• 08:36they differ a little bit.
• 08:38Thanks. And review it too.
• 08:39I learned about this and all the same.
• 08:43Nonetheless.
• 08:46There are functional
• 08:48gradients in the hippocampus,
• 08:50more or less posterior to add heria.
• 08:54Right. So I thought, well let's do this.
• 08:59The only who is very bright post Doc
• 09:02Martin team went out of Melbourne,
• 09:05worked with yeah and Andrew
• 09:072 colleagues of mine.
• 09:09This is a complicated technique that says
• 09:11what's the functional connectivity now?
• 09:13Not within the cortex,
• 09:16between the subcortex.
• 09:18Max, I'm part of the world as the cortex,
• 09:21so we're doing functional connectivity,
• 09:23kind of what I showed before
• 09:25between every hippocampal Vauxhall,
• 09:26or in this case every subcortical Vauxhall,
• 09:29every cortical voxel, each.
• 09:30It's like a you get a Gray,
• 09:33a vector for every hippocampal Vauxhall,
• 09:37then the Navy hearing hammer Vauxhall,
• 09:40we'll have another vector functional
• 09:42connectivity and you just get a similarity.
• 09:45Matrix all subcortical boxes and we get
• 09:50a similarity between adjacent and other
• 09:52hip camp or other subcortical boxes.
• 09:55You do the graphic passing.
• 09:56You just heard about that this
• 09:58morning and you get these lovely
• 10:01looking gradients for this gradient
• 10:03is how quickly is the voxel,
• 10:05the whole brain functional
• 10:07connectivity changing so quite slowly,
• 10:10quite rapidly.
• 10:10That's all we really need to
• 10:12know about these.
• 10:13These are gradients here and Andrew.
• 10:15Use this part of these gradients
• 10:18come up with these beautiful multi
• 10:22style subcortical functional outlets.
• 10:25So when we were chatting about this,
• 10:27we decided, OK,
• 10:29we can do this with the PPI, right?
• 10:32We can see.
• 10:36For these baseline and task PPI coefficients,
• 10:40we'll just do the same thing.
• 10:41So in other words, for every and we
• 10:43just got to look at the hippocampus,
• 10:45every hippocampal Vauxhall better look at
• 10:48the correlation with all cortical voxels.
• 10:52And we look at the next voxel
• 10:54and we'll get 2 vectors and we
• 10:55look at the similarity for that.
• 10:57It looks very complicated.
• 11:00Yeah, and he went back to France.
• 11:03She let him fill the pot.
• 11:04And Kodak could tell you it
• 11:07is actually quite complicated.
• 11:08They're going out on revisions,
• 11:11but at the end of the day I'm
• 11:13going to show you something that
• 11:14says it's not that complicated
• 11:16because it turns out quite simple.
• 11:18So the baseline coefficient,
• 11:21the naive condition, ohh sorry,
• 11:23baseline naive continuing.
• 11:25These are the hippocampal
• 11:27gradients are both of them.
• 11:29This is the first and 2nd principal gradient.
• 11:32There's actually a zero
• 11:33order which is baseline.
• 11:34Both of them show this AP
• 11:37effects one is more or less.
• 11:39To make people and one is anti symmetrical,
• 11:42and those of us who have done a lot of PC
• 11:44I know that most often come in like this.
• 11:47There's a symmetrical effect in the brain.
• 11:50One mode grabs the symmetrical effect.
• 11:53There's some sort of asymmetry
• 11:55left and right aren't saying the
• 11:58asymmetrical mode grabs that effect.
• 12:00These are linearly separable
• 12:01so we add them up.
• 12:02And then there's some other
• 12:04joint modes usually.
• 12:05So one thing that jumps out of this.
• 12:10This is a critical control.
• 12:12We had 33 people with mild AD or MCI.
• 12:17What jumps out of this?
• 12:20Not much difference, right?
• 12:22These these conditions very different.
• 12:25The GLA's were huge,
• 12:27the gradients are all pretty similar,
• 12:29and the gradients are pretty similar
• 12:31in the clinical cohort, however.
• 12:33Ohh. And then the other thing
• 12:35you can do is project these
• 12:37gradients onto the onto the cortex.
• 12:40This is where Punk gets drilled down here.
• 12:43I'm going to talk about this later on,
• 12:45but this is the camel gradient,
• 12:47this is the cortical projection of that
• 12:50anterior posterior Campbell gradient.
• 12:53So not much changes, but what we looked at.
• 12:57Then to do some contrast,
• 12:58we looked at this what I spoke about before,
• 13:01the magnitude of these gradients.
• 13:03So if the gradient changes.
• 13:04Very quickly across a few boxes,
• 13:06we could have a functional
• 13:08boundary there because we say on
• 13:10this side of that rapid gradient,
• 13:11everything's more or less connected
• 13:13to the same cortical voxels.
• 13:15We go down and sort of step
• 13:17and the pattern of whole grain
• 13:19connectivity changes very quickly
• 13:21and then it smooths out again.
• 13:24So we really focused on the
• 13:26magnitude of the of the gradient.
• 13:29This is what the magnitude
• 13:31of the gradient looks like,
• 13:32even though the algorithm was applied
• 13:35to very different sort of data.
• 13:37This is very similar to what we
• 13:39got out of the resting state human
• 13:41Connectome project in the hippocampus.
• 13:43And you can see this step between
• 13:46the posterior hippocampus and
• 13:47the head of the hippocampus.
• 13:49So this is I should have oriented,
• 13:51this is one and here a view
• 13:53of the hippocampus.
• 13:54This is the power of the hippocampus
• 13:55is the head of the hippocampus.
• 13:57You get this sharp.
• 13:59Gradient,
• 14:00which is consistent with what a lot
• 14:02of other people have spoken about,
• 14:03that there is some degree of
• 14:06functional segregation here,
• 14:07but it happens on the background
• 14:09of continuum. So now we can.
• 14:12We can do some permutation tests.
• 14:14We can do some contrasts.
• 14:15We do see differences here.
• 14:18This is the healthy cohort,
• 14:19this is clinical cohort.
• 14:20This is the contrast between
• 14:22the continuing and naive.
• 14:24So even though they're basically the same,
• 14:28these patterns, these gradients,
• 14:29the magnitude of them does change
• 14:32across the task and within the
• 14:34task and across the conditions,
• 14:37we also see slight differences.
• 14:40So this is critical code.
• 14:42Just hopefully control.
• 14:43You don't see much in the resting state.
• 14:46I cause cognitive challenge for the brain.
• 14:49You put in a task and you do see
• 14:51a difference in the functional
• 14:53segregation that you would have is.
• 14:56And that paper?
• 14:59Published today,
• 15:00today in a very famous journal
• 15:04starting with letter end.
• 15:07You're in French?
• 15:11This journal you get the very
• 15:12best reviewers and the best.
• 15:17OK.
• 15:22Now this is where I hope
• 15:24I have a bit of time left.
• 15:27This is where I started to sort
• 15:30of really get into this project.
• 15:35Alright, so this hippocampal gradients
• 15:38is the cortical projections.
• 15:39As I've said, they don't change a lot,
• 15:42not in the magnitude.
• 15:43They do change a little bit.
• 15:45But what's going on here?
• 15:47I started reading about the hippocampus.
• 15:50When I got the reviewers were the only
• 15:52and I got the reviewers the idea was
• 15:55back in Paris with the startup company
• 15:57I was in in Chile in a hotel room
• 16:00with all looking at his Python code.
• 16:03And yeah, this will be this wonderful
• 16:05literature on the campus and the gradient.
• 16:07And it's felt that the panel hip
• 16:10campus is sort of century motor cortex
• 16:13sort of world centric and the head
• 16:16of the hippocampus as you can see
• 16:18here is connected to functionally.
• 16:21Related with key regions and the
• 16:23default plan like curious the angular
• 16:26gyrus anterior called temporal cortex.
• 16:29This is our default mode or what
• 16:32people call the allocentric.
• 16:34It's pretty cool.
• 16:35Like it's like a it's like a Google map.
• 16:39The brain,
• 16:39A1 dimensional Google map of the brain.
• 16:41You know I wanna go into deep
• 16:43like breaks here,
• 16:44go into the head of hippocampus,
• 16:45have all my autobiographical memories.
• 16:48I want to go to entirely with hippocampus.
• 16:50And then they go ohh look I'm in a
• 16:53room speaking to a bunch of people right.
• 16:55So if we in this work by
• 16:58invariant and others,
• 17:00this is like a cognitive map not
• 17:02just a spatial topographic map.
• 17:04It's a cognitive map that somehow links
• 17:08things in the world between time and space,
• 17:10and between what's in the world
• 17:12and what what is in the mind.
• 17:15So what's the meaning of what's in the world?
• 17:19So we can just flip the projection
• 17:22and this is Saturday,
• 17:25the sensory motor and it's pretty,
• 17:28it's a bit messy, it's not beautiful,
• 17:30but these are sort of driving networks.
• 17:32So attentional networks, yes.
• 17:34So this is sort of mapping onto
• 17:37attending the world and the head of
• 17:40the hippocampus is mapping on to.
• 17:42What's going on internally?
• 17:45And this projection, by the way,
• 17:47is quite a lot rougher and clinical
• 17:50in the clinical convoy.
• 17:52OK, so now we're going to tomorrow this.
• 17:56And I'm going to use what's called
• 17:58a neuromas or main fuel model.
• 18:01They're similar but different.
• 18:02So instead of modeling all the
• 18:04neurons or even simplified neurons,
• 18:07I'm going to lump them together
• 18:08and use very simple, great model,
• 18:10as it's called.
• 18:11It's called the Wilson current model,
• 18:13one of the first neuromas models.
• 18:16I had the pleasure to be with me,
• 18:17both Chris Wilson and Howen.
• 18:20And all that is really doing is the
• 18:23right of the red exciter is skills.
• 18:26Going up and then that kicks in
• 18:29and excites the local inhibitory
• 18:32neurons that feedback and inhibit
• 18:35the excitatory cells.
• 18:37That's like a credit credit predator
• 18:40prey and not surprisingly you
• 18:43get oscillations as the exciter
• 18:45is new population are excites
• 18:48the inhibitory population.
• 18:49It is very weird spike phase
• 18:52delay between private selves
• 18:54and the inhibitory cells.
• 18:56And you can then project these
• 18:59rights back into the neurons.
• 19:02And this is what the far
• 19:04and criminal cells reply.
• 19:05And this is what we're firing
• 19:07inhibitory cells like.
• 19:07And we just heard about ripples and
• 19:10these are like fast ripples that
• 19:12sit atop these slow oscillations.
• 19:14And I think this is quite an important thing.
• 19:17It's a lot of time when I
• 19:19present neural mass models,
• 19:20people think it's very abstract.
• 19:22It's just the right modulation of
• 19:24the underlying neurons and you can.
• 19:26Move between two worlds,
• 19:28but the bottom has got
• 19:31thousands of degrees of freedom.
• 19:33The top has got far fewer.
• 19:35And you can see this slide phase
• 19:38lag between the pyramidal cells
• 19:40and the inhibitory cells and
• 19:42this is an importance with
• 19:45neurophysiological recordings.
• 19:46So now I'm going to put these
• 19:49into the hippocampus with a very
• 19:51simple way of coupling just
• 19:54an exponential kernel of the.
• 19:56Local neuron neural masses interacting
• 20:00with their neighbors and it's just
• 20:04going to show this for one reason.
• 20:07This is space prime plot time space
• 20:13space for this particular hippocampal bottle.
• 20:17And this is what it looks like
• 20:19if we put the hip camps here,
• 20:22the posterior and the anterior with
• 20:25this particular semantic kernel,
• 20:27it's more or less going into this
• 20:32seizure like coherent state.
• 20:34But there's a few of these
• 20:37interesting little Birdy bits which
• 20:39are quite interesting to me and if
• 20:42we just focus on them.
• 20:44We see what's happening here is
• 20:46there's a bit of a phase between the
• 20:49excitatory cells and the inhibitory
• 20:52population and then it reorders itself.
• 20:54So it's sort of has a little slip.
• 20:56This is quite common in these
• 20:58sort of systems.
• 20:59I've got this neuron from
• 21:02my friends type of Carmilla.
• 21:04I actually reviewed this paper and I thought,
• 21:06well this is pretty cool.
• 21:07And then working with Matt recently
• 21:10has reinvigorated my interest
• 21:12in computation in the dendrites
• 21:14if the pyramidal cells,
• 21:15and this is just as high published it,
• 21:18come into the distal part of the
• 21:21den tray by the time they get to
• 21:23the soma and the inhibitory cells
• 21:25that are projected down here have
• 21:27had a chance to catch up and that.
• 21:30Cancels the phase delay.
• 21:33Which is this EI balance in the
• 21:36soma until you get this slip,
• 21:39and then you get a bit of excitation.
• 21:41In addition,
• 21:42mismatch and this readout you're on
• 21:45will then fire the readout you're on,
• 21:48Simon,
• 21:48until there's an imbalance between
• 21:51the excitation and inhibition.
• 21:52That's how you get from neural
• 21:55oscillations to some sort
• 21:56of computational spiking.
• 21:57OK, Sir,
• 21:58a few years ago we showed that you
• 22:02can do these neural mask models
• 22:05on structural connectivity,
• 22:07doesn't really matter whether
• 22:08it's the human Connectome project,
• 22:10your local diffusion,
• 22:12reconstructive structural connectome,
• 22:14you get these beautiful whole brain
• 22:17waves and Johanna was speaking about these,
• 22:20but these are really highly
• 22:22nonlinearly excited waves,
• 22:24very metastable and these waves
• 22:26have been published and observed.
• 22:29In many different species,
• 22:32many different recordings and
• 22:34electrophysiological recordings
• 22:35during imaging recordings recently
• 22:38so the not controversial.
• 22:41OK, so now we're going to get waves in
• 22:44the hippocampus and then I'm nearly
• 22:45finished and I'm sorry if I'm over time,
• 22:47but we just changed the synaptic
• 22:50kernel and you can actually hopefully
• 22:52see after a little bit of disorder,
• 22:55you get these diagonal stripes.
• 23:00And if we run out of all moving forward,
• 23:02this is the disorganised part,
• 23:05that of turbulence going on here.
• 23:07And then all the modes competing,
• 23:11they're quite excited.
• 23:12And eventually you get one dominant mode
• 23:15and this is propagating posterior anterior.
• 23:20You might think I'm crazy.
• 23:22I'm not actually.
• 23:23I read this paper first.
• 23:25There's a lot of work coming out
• 23:27now that these alternations be
• 23:29hippocampus travelling waves.
• 23:30So here we are with travelling
• 23:33waves in the canvas.
• 23:35And now we're going to look at the
• 23:37interaction between hippocampal waves
• 23:39and cortical waves in this example.
• 23:41I haven't coupled them.
• 23:43Police are getting you to see
• 23:45hippocampal waves and the cortex
• 23:47doing its own crazy thing.
• 23:49Umm. You get the drift.
• 23:54This is gonna settle down into the waves.
• 23:57This is not.
• 23:58This is just doing its own irregular but
• 24:02quite interesting self organizing patent.
• 24:08But then we're not. We're going to take.
• 24:13These gradients, which I think are phase
• 24:16gradients because you saw that the
• 24:19functional connectivity was quite strong.
• 24:21So I think this anterior posterior
• 24:24gradient is a phase gradient and
• 24:26it maps onto a phase gradient in
• 24:28the cortex and lo and behold,
• 24:31if we just use this map.
• 24:36You're not going to be surprised.
• 24:37Might be surprised. We get.
• 24:44Traveling waves in the hippocampus are
• 24:47reorganizing what's going on in the
• 24:50cortex and imposing a wave pattern.
• 24:52So now we've got waves going from the
• 24:54tail to the head of the hippocampus,
• 24:56driving waves going from the sensory
• 24:58motor to the default mode of the brain,
• 25:01that this is actually A1 dimensional brain.
• 25:03So we'll we'll get there.
• 25:07You see there's a little bit of disorder
• 25:09here, so there's a few times slips, not many.
• 25:11But if I then put a bit more
• 25:15noise into this projection by
• 25:17the Alzheimer's disease cohort,
• 25:19you see already there's a lot more noise,
• 25:22there's a lot more disturbance for
• 25:24the company is not much less cleaner.
• 25:32And. Yes, I had to solve this stick together.
• 25:37I see this is the hope cardboard,
• 25:40and this is the clinical cardboard in black,
• 25:43and the hippocampal waves in red and
• 25:46the cortical waves, and you see.
• 25:50Just a little bit more disorder in the
• 25:54cortex and a lot of disorder in the.
• 25:58In the disease cortex.
• 26:03Now just to finish. Umm, what's happening?
• 26:07Every time it hits and creates one
• 26:09of these spaceships in the cortex,
• 26:12remember, it's reading something out,
• 26:14so there's a lot of false, if you like,
• 26:17read out in the Outsiders disease.
• 26:20There's a lot of more specific
• 26:22readout in the healthy,
• 26:24in the healthy controls.
• 26:25So that's my theory of the brain.
• 26:28But I think it's just like a radar.
• 26:30So if we're walking around daydreaming,
• 26:32we've got all our hipcamp,
• 26:33we've got all our cortical waves
• 26:35doing their crazy.
• 26:35Meeting but if we need
• 26:38to recall recent memory,
• 26:40then we drive the cortex with these
• 26:44hippocampal waves and we're basically
• 26:47using HIP campus to read out recently
• 26:50stored messages in the cortex and
• 26:52then read out through these space
• 26:55slips that then trigger an imbalance
• 26:58in the EI and a little a few spikes
• 27:02that then represent that retrieved.
• 27:06And in our side, this disease,
• 27:09this smooth patterning is disrupted.
• 27:12So you're getting a lot of false readouts.
• 27:14Yeah, that's you.
• 27:16I see.
• 27:23Our work is there and Andrew,
• 27:26everything I showed, all those
• 27:28simulations have done in this financial
• 27:30toolbox which my former post operate.
• 27:33You can download them.
• 27:34You could repeat everything that I
• 27:36did in this talking about 10 minutes.
• 27:39This is my fledgling group in Newcastle.
• 27:42I'm beautiful and wonderful
• 27:44country and it next time you're in
• 27:46Australia you must come and visit.
• 27:48We can go surfing.
• 27:49It's going snowboarding.
• 27:51OK. Thank you.
• 27:58Randy. Interestingly,
• 28:01the surfing connection package
• 28:03looking at flowing waves.
• 28:05Nice connection there.
• 28:08There is, actually, because
• 28:10when you're looking at whites.
• 28:11It was talking to me, Mac.
• 28:15No borders, right,
• 28:16they're looking ahead. No, not me.
• 28:18I've got about A2 metre window,
• 28:21but you're looking further ahead now.
• 28:22Wave good surface and the same thing.
• 28:25They can read the wave 3040 metres
• 28:27away because they know the reef.
• 28:29When the wind is coming,
• 28:30when ocean waves coming and they interact
• 28:33with the underlying ocean floor,
• 28:35it causes disturbances.
• 28:36And that's where this whole
• 28:38crazy theory of mind came from
• 28:41here the campus is getting the
• 28:43brain to scan from back to front,
• 28:45from sensory motor to default and back.
• 28:48And all the little disturbances
• 28:49that were there from a few minutes
• 28:51ago caused disruptions in the
• 28:53I balance and get read back,
• 28:55back into the back, into the hologram.
• 28:57So sorry right here.
• 29:02Wait to be phased?
• 29:05So that makes it more clear with
• 29:07the theory is I guess the question
• 29:09I have is the degree to which the
• 29:12phenomena you're you're modeling is
• 29:14self organization versus actually
• 29:16more directed this comes to mind
• 29:18is our little man sort of pushing
• 29:20it tends to do that or does this?
• 29:24Organization emerge naturally
• 29:25in this always happening by by
• 29:27the nature of the anatomy of the
• 29:29function that are that couple.
• 29:30Well the key thing is in the hippocampus.
• 29:33The key ingredient I made out of hippocampus
• 29:36have a slightly slower time scale,
• 29:37faster than the head of the hippocampus.
• 29:40But naturally after some disorganization
• 29:43self organizes into these coherent waves.
• 29:46The brain is doing its own thing,
• 29:48the cortex is doing their own thing
• 29:50until you turn the coupling on
• 29:51and they're just like the atrium
• 29:53driving the ventricle of the heart.
• 29:55The hippocampus drives the cortical waves,
• 29:59reorganizes them so that they
• 30:01flow from sensory motor cortex
• 30:03through to the default mode.
• 30:05So it is self organising and you just
• 30:07need a bit of symmetry breaking and
• 30:09then you can stand back and it all comes
• 30:12out without being heavily directed.
• 30:15And it comes back to like Jim was talking
• 30:18about the the learning, what he called it.
• 30:24You're essentially modeling what a
• 30:26complex system adaptive system do
• 30:27after go through iterations of has
• 30:29this capacity to explore repertoire
• 30:32to use the word dialect and you
• 30:34by by putting the model in there
• 30:36as it as it moves through time as
• 30:38the model grows becomes coupled.
• 30:40It has this natural capacity to
• 30:42explore and the exploration of lighter
• 30:44stand your models is driven by at
• 30:46least coordinated through the caps.
• 30:48That's the practical importance.
• 30:51Turning on the right on, yeah,
• 30:53there's quite a lot of theory that
• 30:56recent memories are encoded by latent
• 30:58excitation in prefrontal cortex.
• 31:00And when the hippocampus
• 31:01sends the cortical wave,
• 31:03just like my surfer on the reef,
• 31:05it disrupts the wave and
• 31:07it disrupts the phase.
• 31:08Relationships would be hidden campus that
• 31:10causes a disruption in the hippocampus.
• 31:14Yeah, it's. So it's more about scanning.
• 31:16It's not, this is not a brain exploring
• 31:18when the hippocampus gets turned off.
• 31:20That's the sort of daydreaming mode.
• 31:22But I will say with Tim's talk
• 31:25for where is Tim is, yeah, look,
• 31:27the activity is more or less the same.
• 31:30It's the various relationships
• 31:32amongst the Pastor networks that are
• 31:35doing everything that's different
• 31:37from daydreaming to cognition.
• 31:39And I think you were saying the same too,
• 31:41Todd, the activity atop the resting.
• 31:43So it's not that different, but yeah,
• 31:47these patterns are different.
• 31:49Your honor.
• 31:52Be nice.
• 31:56And actually there is a study where there
• 31:59are doing like temporal interference,
• 32:02you know with our simulation content frontal,
• 32:05Netherland, posterior part of the campus.
• 32:10And they find that modulates like
• 32:13sending their different waves.
• 32:16In different parts of this campus
• 32:19modulation different networks
• 32:20so and in a way also like slide
• 32:23change of frequency but if the
• 32:25they were comes descending like a
• 32:27traveling with the rain and then
• 32:29will resonate inside of the brain
• 32:31stretcher and generated this campus.
• 32:33I think it's a I I really buy this as a as a.
• 32:38Then there are other structures also
• 32:40sending different types of sonars
• 32:42or radars like you're saying for
• 32:44what type of wave is being sent?
• 32:46Maybe there are more strangers sites.
• 32:50It is the alpha, right?
• 32:53That's the, that's the,
• 32:54that's what Hansberger described
• 32:56and we've worked on that before.
• 32:57That's a different resonant loop.
• 32:59The hippocampus is more feeder and
• 33:01then the ripples are top of that.
• 33:02Yeah. So yeah,
• 33:04one physicist to the other, yeah.