What happens if there is no brain plasticity

[fredreload]

What would happen if there is no brain plasticity? Would the brain be incapable of gaining new memory? Say you grow to a certain age, obtain enough memory, feeling, and such then the brain plasticity stops(yes, I know it is not supposed to), what would happen to that person? Is he still able to perform like a normal individual?

 

[DaveC426913]

I think that requires a nuanced definition of plasticity. Does it include forming new memories? If so, then you'd essentially get the plot of '50 First Dates'.

 

[DiracPool]

What would happen if there is no brain plasticity? Would the brain be incapable of gaining new memory? Say you grow to a certain age, obtain enough memory, feeling, and such then the brain plasticity stops(yes, I know it is not supposed to), what would happen to that person? Is he still able to perform like a normal individual?

Well, I think you answered your own question. The answer is NO, you can't form new memories without brain plasticity. The word "forming" implies change. You can't reform anything absent plasticity. It's the plasticity of the fiber connections between neurons in the cortex that form memories. We neuroscientists call these "Hebbian connections."

 

[jim mcnamara]

There are people who - due to trauma or disease - have limited or no short term memory. In general, they cannot live alone or unassisted - safely. So, the neural plasticity you talked about is important to everyday living.

 

[fredreload]

How about anyway to induce brain plasticity? Excite the external of the neuron to create new memory path using electromagnetic radiation or chemical to get the correct memory structure?

 

[DiracPool]

How about anyway to induce brain plasticity? Excite the external of the neuron to create new memory path using electromagnetic radiation or chemical to get the correct memory structure?

https://en.wikipedia.org/wiki/Hebbian_theory

 

[fredreload]

Hmm, kind of a bummer they can't train artificial neurons to perform neural plasticity. You think they have a way to bypass that? I was originally thinking of a brain structure made of 3d printing with a different type of material to substitute for neurons, but it wouldn't work unless it can do neural plasticity.

 

[DiracPool]

I was originally thinking of a brain structure made of 3d printing with a different type of material to substitute for neurons, but it wouldn't work unless it can do neural plasticity.

A different type of material? Like what? Do you really think we can "bake" up a chemical compound that functions like a neuron and put it in a 3D printer? It's not that easy.

Hmm, kind of a bummer they can't train artificial neurons to perform neural plasticity. You think they have a way to bypass that?

It's not difficult at all to do this. However, modeling a single neuron is not the level of modeling that you want to focus on. You want to focus on the mesoscopic level of the population that we model as coupled oscillators. You can model on the order of 10,000 neurons as a "node" and use one parameter as a "weight" to simulate the modifiable synapse. We can simulate these using non-linear coupled ODE's. Again, it's not that difficult to simulate these oscillators in electronic circuits.

http://sulcus.berkeley.edu/wjf/DF_Principe_IEEEAnalogVLSI.pdf

We've found that the limiting factor is in the complexity of the scaling factor of the oscillators. The actual biological brain is a 3D structure which allows for a rich and complex interconnectivity. Trying to emulate that 3D architecture with electronic circuitry is the current challenge we face.

 

[fredreload]

A different type of material? Like what? Do you really think we can "bake" up a chemical compound that functions like a neuron and put it in a 3D printer? It's not that easy.
It's not difficult at all to do this. However, modeling a single neuron is not the level of modeling that you want to focus on. You want to focus on the mesoscopic level of the population that we model as coupled oscillators. You can model on the order of 10,000 neurons as a "node" and use one parameter as a "weight" to simulate the modifiable synapse. We can simulate these using non-linear coupled ODE's. Again, it's not that difficult to simulate these oscillators in electronic circuits.

http://sulcus.berkeley.edu/wjf/DF_Principe_IEEEAnalogVLSI.pdf

We've found that the limiting factor is in the complexity of the scaling factor of the oscillators. The actual biological brain is a 3D structure which allows for a rich and complex interconnectivity. Trying to emulate that 3D architecture with electronic circuitry is the current challenge we face.

A different type of material like carbon nanotube that can be 3d printed into a neuron shape, it doesn't have to be a real neuron, as long as it runs electrical synapses it should work as a real neuron, it doesn't need to be alive. Well how you go about creating neural plasticity for a carbon nanotube brain structure is where I got lose. My idea is substituting real neuron with carbon nanotube because if you 3d print a brain with real neurons, a lot of things can go wrong, but if you 3d print just the structure and substitute it with a more durable material then it should work better, again it is never tested. So if you would let me know how to create neural plasticity for carbon nanotube or any other material it would be cool. The problem is I think neural plasticity is built into the internal structure of the neuron so that feature can't really be transferred to a carbon nanotube.

 

[DiracPool]

So if you would let me know how to create neural plasticity for carbon nanotube or any other material it would be cool.

That's not going to happen. A neuron is a very complex system with a lot of moving parts. A carbon nanotube comes just as advertised, it's just carbon. You're not going to get any functionality out of it other than structural.

 

[atyy]

That's not going to happen. A neuron is a very complex system with a lot of moving parts. A carbon nanotube comes just as advertised, it's just carbon. You're not going to get any functionality out of it other than structural.

How about:
http://www.research.ibm.com/articles/brain-chip.shtml
"We envision augmenting our neurosynaptic cores with http://www.modha.org/papers/013.CICC2.pdf to create a new generation of field-adaptable neurosynaptic computers capable of online learning."

 

[fredreload]

Hmm, we'll need some type of material that can be updated with structural changes to simulate brain plasticity. Maybe some type of nano machines would do the work? This sounds like a hard problem.

Hardware sounds like a possible answer. Devices like hard drive can easily changes their memory.

 

[fredreload]

How about:
http://www.research.ibm.com/articles/brain-chip.shtml
"We envision augmenting our neurosynaptic cores with http://www.modha.org/papers/013.CICC2.pdf to create a new generation of field-adaptable neurosynaptic computers capable of online learning."

Is it water proofed?

 

[madness]

I think the consequences would be pretty severe. At least some types of plasticity (https://en.wikipedia.org/wiki/Homeostatic_plasticity) are required just to remain stability. Without this, the network would be likely to end up in either an epileptic or braindead state, or at the very least a state that is far from its operating point.

 

[fredreload]

Not if we can create a virtual brain with neural plasticity and link that to the real brain

 

[atyy]

http://www.nature.com/scitable/blog/scibytes/hippocampal_prosthesis_do_scientists_have
(I'm not sure how far along this is, and whether the current implementations have synaptic plasticity.)

 

[madness]

Not if we can create a virtual brain with neural plasticity and link that to the real brain

Was this a response to my post? I'm not entirely sure how what you are proposing would solve the issues of stability that arise when homeostatic plasticity is removed. It may be possible to artificially inject current into a neuron in order to keep the input statistics fixed over time, rather than adjusting the way the neuron responds to those inputs via homeostasis. That's not really the same as adding a virtual brain with plasticity though, it's just adding some kind of control system.

 

[fredreload]

Well, we'll have to agree that a real brain can link to a 3d brain inside a computer, theoretically

 

[madness]

Well, we'll have to agree that a real brain can link to a 3d brain inside a computer, theoretically

I'm not clear on what that link would entail. Having a conversation involves linking two brains in some sense. My two brain hemispheres can be considered as two linked brains in some sense, as can any arbitrary partition of my brain into two subsystems. Whether a real brain can be linked to a brain inside a computer in some meaningful sense I'm not sure.

 

[fredreload]

Ya, it's quite an interesting idea, my guess is it works, we just need a way to test it outP.S. And if I am wrong, nothing to lose, another reason I guess it would work is because the artificial brain would also contain a consciousness