that's not particularly relevant here. All of this investment and technology is based on reading out motor cortical activity intended for controlling muscles. We're FAR away from decoding thoughts, whatever that even means.
Although Google is usually the brand associated with self-driving cars, DARPA actually did most of the work. The self-driving car challenge went on for many many years with ever increasing goals until DARPA decided the tech works now and stopped the competition.
Google then hired the team that won the last several competitions and kept the ball rolling.
Seems to me like this is the only path towards true innovation since companies are great at exploiting existing tech and making incremental and efficiency improvements but are usually really bad at making great leaps forward. If you provide incentives divorced from the market for desirable goals the market can later pick up the successes and build on them.
Or computers. The military threw a lot of money in to early computer research in order to make nuclear weapons work (better), cryptography, ballistics tables, and other military purposes.
I'm involved with one of these programs and am very encouraged to see DARPA's focus as very much inline with the goals of the broader field of neuroscience.
In heading off some of the anticipated snark, this is very much a program with the goal of understanding basic science and developing tools to better understand how our brains work.
Anyone who wants to see the dramatic impact of deep brain stimulation needs to see this video[1]. Just by turning off and on his neurostimulator (see starting 2:00-ish) see the dramatic change in his body.
To clarify the above, the video starts with the deep brain stimulation device turned on. The changes mentioned are due to the Parkinson's Disease effect showing themselves once the DBS device is turned off.
That's a really interesting video on this debilitating disease. Are there any notable examples of where DBS has failed or what could entail as "the risk" of undergoing such a surgery?
I have a friend whom I can refer about this procedure.
DBS surgeries are somewhat invasive as they involve sticking a 3mm electrode all the way through cortex and deeper structures into the basal ganglia. That being said, anecdotally I believe the side effects are mostly modest, though can't speak on that authoritatively.
it's actually not too bad. There is a small micro electrode that is sub-millimeter, and the DBS electrode itself is ~1-1.5 mm depending on the model.
The one traumatic part of it is that some groups use a device popularized by some of the French groups that use a 5-microelectrode array for the mapping portion of the surgery, which has a 4x4 cross sectional spread. This probably pins and destroys a lot of the brain tissue in its path, but the people who use it claim it gives them "better spatial resolution" in the mapping recordings. Which doesn't apparently improve overall accuracy anyway, but some people are married to using that damn device.
Most surgeons in the US use 1-3 single electrode passes sequentially, which is a lot better.
Truth be told, these are early days even though we have used stimulation for a while now (artificial cardiac pacemakers). After all, this is brain surgery. You implant electrodes bilaterally into the deep brain (thalamic region for Parkinson's) and send small shocks. The hope is this in turn will impact the neuronal signaling that causes excessive or too little movement. In addition, it is not clear what the signals do the nearby neurons (question: do they die and in turn what happens to other brain functions etc.) and impact over longer term as you keep stimulating (what happens to the neurotransmitter responses etc.). So, definitely a lot of studies like the DARPA one need to be undertaken.
So the usual caveats apply. You need to have tried other means (pharmacological, exercise), need to be healthy and willing to risk adverse consequences. In the case of the video linked, it is somewhat of a no brainer. The person is young 39 years old with a family. He seems to have a pretty bad case of Parkinsons. I'm sure he is willing to enjoy a much better quality of life even if it is risky.
Actually, DBS has been around since the 1980's, just Medtronic sat on the patent and it was little more than a glorified converted cardiac pacemaker. now the patent is up, all of the sudden all sorts of research into closed-loop stimulation, electrode designs, battery design, etc suddenly pops out of the woodwork.
They're trying to do some sort of closed loop feedback instead of some clinician inputting random settings of volts/frequency/pulse width into the device, but no one knows exactly what signals in the brain to listen for yet. The EE-oriented neuro types are looking at signal processing, i.e. using FFT's to look at power spectra of local recordings, etc, but nothing too promising yet.
As far as what the physiologic mechanism of the electrical stimulation does, no one really knows, but there's some data from monkeys and mice optogenetics suggesting it activates passing axonal fibers. Generally, nerves don't die due to the electricity, unless your surgeon does the implant French-style and shoves a 5-microelectrode array into your head, therefore boring a nice 2-4mm tunnel into your brain...
Typical risks - there's a 1-2% prevalence of hemorrhage which usually isn't too bad. It's really rare to have a major event. The other major risk is infection (as with all implanted hardware), or device malfunction.
Otherwise, the most common side effects that could happen is that the stimulation causes some spasms or tingling. There's some hand-waving as to whether it may increase susceptibility to depression or falls, but that is hard to tease out between disease progression.
Of note, if it's Parkinson's, one hard part is that the meds will have to be adjusted a lot post-op which cause some folks problems.
It's pretty terrifying what the implications of this could be. Want super soldiers who will obey any order given? I imagine they'll try to experiment with this kind of control mechanism.
On the inevitability human computer interfaces leading to some form of Übermenschine: "We just have to hope that Dick Cheney isn't the first person they plug in."
We need to monitor thoughts of people to prevent them from committing crime. We promise* not to use that information in any* bad* way.
* For a very specific meanings of promise, any, and bad which are all unfortunately classified.
But seriously, this research can enable us to understand brain better and help a lot of people.
A lot of this research is going towards developing neuro-prosthetics. The primary application of neuro-prosthetics in humans is to give paraplegics either the ability to control a cursor with their mind (a huge improvement in standard of living) OR the ability to control a prosthetic arm or leg with their mind (the long term goal).
Currently, this is only available with invasive brain surgery that can often have complications. So money spent on better imaging and implant technologies will have a strong positive impact on the field.
Interestingly, the researchers I know in this area are confused about why such a big deal is being made about this "Brain Initiative" because the amount ($70M) is actually not a lot given how capital intensive this type of research is and how many labs it will be spread amongst. Still, any funding is better than no funding.
That is true - I don't know how many grants/VC rounds a company like Braingate has blown through. The Andersen group seems to have success as well, maybe with less actual expenditures.
I'm at Stanford's neural prosthetic systems lab which has previously been funded largely under the DARPA REPAIR program.
Minor point of clarification - while the surgeries to implant the electrode arrays are technically a neurosurgery, they're minimally invasive in the sense that it doesn't actually disturb or destroy any brain tissue since the motor cortex is accessible from the outer surface. The primary concerns now are not complications with surgery, but performance of the electrodes over timescales required by patients (several years minimum).
I'm not aware of any significant complications in any of the BrainGate clinical trial participants to date (if there have been, I'm not aware of them and would be a little surprised to hear it).
>they're minimally invasive in the sense that it doesn't actually disturb or destroy any brain tissue since the motor cortex is accessible from the outer surface.
There's a slogan, "you're never the same after the air hits your brain". Is that hyperbole then?
The neuroscience community in general has had mixed reactions to the Brain initiative during the period where we haven't had much information. It was initially announced as something much larger, then pared back, and it wasn't clear whether it would come from existing funds or from newly allocated funds, which was the main question most researchers had.
Google glass is not a moonshot. Google brain implant will be - just think about the SAT scores and all your collect test scores difference for those with brain implant and those without.
But on the downside, their every waking thought will be uploaded to Google++ and YouBrain, and every now and then some part of their knowledge base will be 'discontinued'
41 comments
[ 0.24 ms ] story [ 73.7 ms ] threadIs there an equivalent of a tin foil hat that is available for subdermal implantation?
Kidding aside, military research has resulted in some of the most amazing stuff these days... like the entire space program and velcro.
Google then hired the team that won the last several competitions and kept the ball rolling.
Seems to me like this is the only path towards true innovation since companies are great at exploiting existing tech and making incremental and efficiency improvements but are usually really bad at making great leaps forward. If you provide incentives divorced from the market for desirable goals the market can later pick up the successes and build on them.
In heading off some of the anticipated snark, this is very much a program with the goal of understanding basic science and developing tools to better understand how our brains work.
[1]http://www.youtube.com/watch?v=uBh2LxTW0s0
I have a friend whom I can refer about this procedure.
http://www.mayoclinic.com/health/deep-brain-stimulation/MY00...
The one traumatic part of it is that some groups use a device popularized by some of the French groups that use a 5-microelectrode array for the mapping portion of the surgery, which has a 4x4 cross sectional spread. This probably pins and destroys a lot of the brain tissue in its path, but the people who use it claim it gives them "better spatial resolution" in the mapping recordings. Which doesn't apparently improve overall accuracy anyway, but some people are married to using that damn device.
Most surgeons in the US use 1-3 single electrode passes sequentially, which is a lot better.
So the usual caveats apply. You need to have tried other means (pharmacological, exercise), need to be healthy and willing to risk adverse consequences. In the case of the video linked, it is somewhat of a no brainer. The person is young 39 years old with a family. He seems to have a pretty bad case of Parkinsons. I'm sure he is willing to enjoy a much better quality of life even if it is risky.
They're trying to do some sort of closed loop feedback instead of some clinician inputting random settings of volts/frequency/pulse width into the device, but no one knows exactly what signals in the brain to listen for yet. The EE-oriented neuro types are looking at signal processing, i.e. using FFT's to look at power spectra of local recordings, etc, but nothing too promising yet.
As far as what the physiologic mechanism of the electrical stimulation does, no one really knows, but there's some data from monkeys and mice optogenetics suggesting it activates passing axonal fibers. Generally, nerves don't die due to the electricity, unless your surgeon does the implant French-style and shoves a 5-microelectrode array into your head, therefore boring a nice 2-4mm tunnel into your brain...
Otherwise, the most common side effects that could happen is that the stimulation causes some spasms or tingling. There's some hand-waving as to whether it may increase susceptibility to depression or falls, but that is hard to tease out between disease progression.
Of note, if it's Parkinson's, one hard part is that the meds will have to be adjusted a lot post-op which cause some folks problems.
But seriously, this research can enable us to understand brain better and help a lot of people.
Currently, this is only available with invasive brain surgery that can often have complications. So money spent on better imaging and implant technologies will have a strong positive impact on the field.
Interestingly, the researchers I know in this area are confused about why such a big deal is being made about this "Brain Initiative" because the amount ($70M) is actually not a lot given how capital intensive this type of research is and how many labs it will be spread amongst. Still, any funding is better than no funding.
Minor point of clarification - while the surgeries to implant the electrode arrays are technically a neurosurgery, they're minimally invasive in the sense that it doesn't actually disturb or destroy any brain tissue since the motor cortex is accessible from the outer surface. The primary concerns now are not complications with surgery, but performance of the electrodes over timescales required by patients (several years minimum).
I'm not aware of any significant complications in any of the BrainGate clinical trial participants to date (if there have been, I'm not aware of them and would be a little surprised to hear it).
There's a slogan, "you're never the same after the air hits your brain". Is that hyperbole then?
70 mn. over five years is (naively) 14 mn. per year -- which is 0.05% of DARPA's budget.
Google glass is not a moonshot. Google brain implant will be - just think about the SAT scores and all your collect test scores difference for those with brain implant and those without.