1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
|
Our group is a neuroengineering group. There's an active volunteer program, where there's half-a-dozen people initiating projects. Some of them have initial origins in retirees or or the part of the head. There's unrestricted research, we give people who don't even donate, there's a nice feedback loop that we're excited about.
The tools allow us to understand the brain at a level to understand it. The brain over the last 100 years has been about looking about the molecular scale and so on. There is also an importance about the diseases related to atrophy, poluttion, and different cells or losses, like epilepsy or disorders. One of the most key ideas is that we can do a new generation of treatments that are about targeted neuromodulations. You can target the circuits in the brain, and there's this immeineslsy complicated 3D microcircuit structure. The trends are pointing towards this being increasingly common. More than 100k people have invasive cochlear implants. More than 50k have deep brain stimulators. There's even in adolescents, not even the traditional population. And so on.
So, why control neural activity? We can reprogram the neural computations that underly disorders. If we just give someone a drug, it will bathe the brain in a substance, and that cause problems. The brain is a very cryptic thing. We want to target information in a targeted area in the brain. There's driving activity in technology, so one of the technologies is the little spikes or action potentials in the brain. So, light's nice because you can point at a population in the brain, or selectively. This is useful.
The experiment begins a decade ago where we were trying to find genomic materials for photosynthetic pathways, to convert light to electrical energy, we put them into cells with viruses, and we make cells with a light. Here's a cell that was coated with the sensors, like coating with solar cells. There's all sorts of reactions happening. We can put them into the brain. One of the big things is trying to do this with 3D optical fiber arrays to beam the light in and do patterns. They can do complex and distributed structures. This is doing the hippocampus; it was compromising the ability to form new memories.
One of the examples that we're doing here, one of the PTSD grant, from the DARPA. Can we find targets in the brain that can ameliorate the pathological fear of events that were previously neutral? We're going back to Pavlov and so on, like operant conditioning and pavlovian conditioning, like enducing fear states using a tone. WE're trying to figure out which ones can amiliorate the different sites in the brain, and then we can find which targets might not treat the disorder.
The other thing that we've doing a lot of is new genomic diversity, so that we can mine it for better things. Nature is good at making things. Maybe we can team up with nature, and find interesting stuff, put them into biological systems, and treat them. Insulin, taxixiciln, you name it. Earlier this year, we had this whole diversity of organisms from all over the planet, and look at the genes, express them into neurons so that we can do other things. We can completely turn off the neurons with one method. This was one of the methods to look at more and more ecology populations and samples, and actually came out and get out some scuba diving. We want to mine the genomic diversity, so that we can connect genotype, and the amount of genome data that we're doing. The genotype that we're doing is not at moore's law rate yet, so we can cure stuff.
There's different classes of models in the wild. One of the proteins is sensitive to blue light, some is sensitive to other colors of light. If we can only figure out how to phenotype it quickly. And the ability to cure diseases. A lot of our work is based on more collaborations..
Which is where I want to start thinking about brain coprocessors. If you look bad at the opening slides, we talked about cochlear implants. If you can mine the data, and then use intelligence, and there's two multi-million dollar projects to merge AI with neuroscience, get the information where it needs to happen, and have maximal clinical benefit in the short term. We can figure out where natural intelligence and ai starts to happen. We can train the software on neural computations in the brain, and that's speculative and we haven't figured out how that would work.
The other interesting thing that could happen with this is that the technologies could be used directly - proteins and light - as we expand the genomic and ecological search- if you think about it, gene therapy, which is required to put its genes into cells. One of the things about this is that there are classes of viruses, adenoviruses, lentivurses, that are very safe to use in humans. Adeno-associated viruses (AAV), FDA trials, and there has been no single adverse; there's payload issues sometimes, of course. It's quite inocculouous. One of the things that we did is a pre-clinical study, and with two groups, Bob Desimone, Ann Graybiel. One of the things we did was pre-clinical studies with inocculant parties. So we gave them some of the molecules, and this was the gold standard for neurological psychiatric standing, there was this pre-clinical; we found we can drive neurons over a long period of time. We worked with phenomenologists, for antibodies, and we found no evidence for immune reactions. These are molecules from algae and so on, one of the things we're exploring is how to use these things.
We're collaborating at USC; there's a company he's funded to see if we can do neural disorders- like blindness. And classical disorders are due to loss of photoreceptors, and they are here, gone here. Why can't we just make this back? Why can't the retina produce the photoreceptors again? one of the things that we've done is taken a model of blindness, and the kind of mice that you find in pet stores, and these are similar to the kind of things that happen in humans. And these are going down in long alleys, and over many weeks of training, they come up with strategies, like don't go the same pathway twice. Untreated rd16. Here's a mouse that was blind a few weeks before, and received our light receptors in a single dose, and he can solve the maze just as well as identical rats that could see all their life. That's what we've been up to.
|
