Friday, May 30, 2008

More on Schwartz lab paper

Um, whoops there. Not true. Our lab did it before. Here's the video in the supplemental info from the 2006 paper. Granted that Andy's robot is more impressive.

Also, notice that one of the Nature videos for the monkey-bot paper covers the arm, while the others show the arm movement. Compare this to this. They're both clearly shot for mass consumption, because view of the headstage is obviously being obscured. If it's not a big deal, why hide it?

And finally, the supplemental info reports everything in much more detail. Like that the two monkeys had completely different electrodes (one array, one microwire), so that the number of cells used was significantly different (>100 vs <30), and it took a month of training. The microwire monkey had slightly worse performance (78% vs 61%).

There was also an NPR interview where he makes a couple false statements, and mentions doing stimulation of somatosensory cortex in future human implants, even though he has been highly critical of any use of the technology in humans. I don't understand the 'tude, dude.

Wednesday, May 28, 2008

Monkey self feeding BCI control

Alright, I get it. Several people have sent along links to Andy Schwartz's Nature publication, so here's the info. Thanks to Remy Wahnoun and Natalia "Get Moose and Squirrel" Bilenko for doing my work for me and compiling the links, and being the first to send info along.



http://www.nature.com/nature/journal/vaop/ncurrent/full/nature06996.html
http://news.bbc.co.uk/2/hi/science/nature/7423184.stm
http://technology.newscientist.com/channel/tech/dn14000-robomonkeys-use-brain-power-to-grab-a-bite.html?feedId=online-news_rss20

The work reported involves real time decoding to drive a robot arm, which fed the monkey. I haven't had time to run through the whole thing, so here's the "shoot from the hip" version of my thoughts. I'll update these

Interesting points:

- The monkey was trained to use the arm with a joystick, and followed by, the part I thought was very interesting, various levels of pre-programmed control assistance while using brain signals. I will expand on this later, but it is an interesting task design for several reasons.

- Having seen some of the raw video at the Neurobotics 2007 workshop, there is much more obviously correlated arm motion. They have an entire, substantial paragraph devoted to this, since they probably knew it would be a criticism.

The three points they make are:
1) Movements were with the arm ipsilateral to the electrodes.
2) The movements were delayed by up to a second.
3) Moving is required, since BCI control has been demonstrated before without it.
They cite some supplementary info and write, "... monkey’s hand movement was only loosely coupled to prosthetic control." The video I saw before tells a slightly different story. There was a pulling movement 100% of the time (every trial), and when the arm didn't reach the monkey's mouth it would 'paw' or 'dig' (repeated movements similar to pulling an invisible object toward the body) repeatedly until the arm reached the monkey's mouth. You can see this pretty clearly in the video linked above (watch the hand in the plexiglass tube). Yes, there is a delay, but the whole situation would be interesting to examine. It may not require the 'pulling' at onset, but monkeys very often use both hands to eat. Those of you with monkeys, go give them a treat. They will grab it with they right hand (usually), and then pull it towards their body and engage their left hand as they do so. Feeding it almost always a bimanual task for macaques, explaining the ipsilateral activation and the delay. Reason #3 above is kind of a brush off reason - hey, movement is needed to use a BCI, so no big deal. The three cited papers are all in humans, all were tetraplegic, all lacked some degree of propriocentive feedback, two didn't use PVA or multielectrode arrays, none used an arm, none had confirmed the lack of covert movements with EMG, and the list goes on.

- The trial was very organic and continuous, which is great. From my brief skimming, it looks like simultaneous control of the hand aperture needed to be maintained throughout the task, hinted at when they observe the the hand slowly opening along the path to the food.

In general this is a nice, brief summary with some very interesting points (perfect for a Nature Letter). Obviously they are preparing some studies, and this was released to wet out appetites. The 'organic' nature of the task will open it up to many criticisms, but more studies like this are needed in order to understand the dynamics in play when dealing with a experiences that lack built-in boundaries and environmental awareness. Definitely good thought fodder.

Monday, May 26, 2008

Fox News - Accurete Repoting!

This, what I guess passes for an article at Fox News, mentions 'The Lab' along with my old 'The Lab'. The quotes, images and, well, all the content regarding the current 'The Lab' is straight out of video clips or other publicly available material. I think they were just going to write an update on Phil Kennedy's work and the reporter needed to add some fluff. So, go ahead and take a peak if you want. You might want your 5 minutes of life back, though.

In REAL news, can't say anything more, but you might see me in the background on a upcoming major news show airing in the Fall. not saying anything more until everything is public, but that's part of the reason for the slowed updates lately.

And just a quickie on the first topic. I find it odd that every 6 months or so, there is an update on the project here. I mean, yes, we make progress, but the updates seem to come randomly. Sometimes they're started by a rogue Reddit or Digg post, sometimes there's a big media blitz. I know that there are times when the company is looking for investors and when we get publications, where I expect media coverage, but at least half the stories I've come across have been completely out of the blue. The only possible conclusion I can come to is that people are weird. That is all. If you read the Fox article and read this, I bet you want your 10 minutes of life back. Too bad, they're mine now.

Tuesday, May 20, 2008

Complexity

Just wanted to point out the MIT journal Artificial Life has some nice looking articles on complexity, emergence, and evolution in the current issue. Looks like some good reading.

Friday, May 16, 2008

Amputees in the Olympics


Two big stories today.

First, Pistorius has won his appeal, and will be allowed to run in Beijing. Pistorius is a double amputee mentioned in previous posts here.

Second, a swimmer who had their leg amputated below the knee qualified for the Olympics this year, making her the first in history. Oh, and no wimpy prosthetics, either. She did it with one leg.

Major news

Distancing myself a little here, but this isn't unexpected at all:

Form 10-Q for CYBERKINETICS NEUROTECHNOLOGY SYSTEMS, INC.

"On May 15, 2008, we entered into an Asset Purchase Agreement (the "APA") with I2S Micro Implantable Systems Inc. ("I2S" or "Purchaser"), a Utah corporation located in Salt Lake City, Utah. The transaction is scheduled to close on or about May 16, 2008 pending execution of certain ancillary agreements."

Thursday, May 15, 2008

Cyborg art

A weekend cyborg art exhibition is going on in New York. Just FYI - catching these stories gets you names of artists interested in human-tech integration, which = new presentation intro slides and wallpapers for your PC. Behold Cyberdine (via IO9).

And after writing that, I thought, "Hey... I got some more related stories..."

- Ancient medical illustrations from the mysterious 'Orient' have been placed online, along with 3D models of early wax medical models.

- A vibrating interface for a blind artist (I swear I posted this before, but it was still flagged).

- On the art of wordsmithing - influential scientists divulge the books that inspired them.

- EEG controlled robots as part of another art exhibit.

- A similar neuro-art rundown was posted a while back on Developing Intelligence.

Wednesday, May 14, 2008

Death to the spike sorters!

First, good paper here: Spike train decoding without spike sorting by Ventura, V. (also available free on her site here)

Essentially, she looks at using a simple voltage thresholding of the continuous data stream to infer neuron identities using a very complicated process of turning curve fitting and statistics. One of the more interesting things, I thought, was the way in which noise drops out of the equation, since "noise" is all allocated to a "noise neuron". I am just looking at the paper for the first time, but I swear I remember her stopping by my SfN poster last year (we did a similar spike-sorting-less type of approach as on of our signal sources for comparison and it did perform quite well), and exclaiming something like, "I knew I wasn't the only one that thought this should work!"

If you're wondering what I'm talking about, the idea is this: You get a signal from an electrode channel that is the combination of many different types of "field potentials" and "unit activity". Unit activity is easy enough - those spike looking things that people use to discretely decode something or other. The tick marks in a raster, the counts in a histogram. "Field potentials" are more complicated, and represent the total electrical activity that the particular electrode is being subjected to and able to convey, based on its materials and geometry (thing like impedance, etc.). This represents things like the slow fluctuations in membrane potential of nearly cells as ions move through channels, overlaid activity from cells within the volume of brain matter, depending on location and electrode sensitivity it can include EMG and EOG signals, etc. Essentially everything that has an electrical charge big enough and near enough to be detected.

When you use unit activity to decode some type of behavior, you have to decide how many neurons you are recording from, or so the thinking has been. Excitatory signals spike, as do inhibitory, so people generally look at each different waveform as a separate cell's activity. In order to separate the waveforms, you "spike sort" (or "cluster cut" if you're an old DataWave user). This involves various ways of saying "this waveform is different from that waveform". It is much more difficult than you think, and is a major pain in the ass for a number of reasons. It also adds a ton of processing overhead, making it particularly annoying for BCIs, since you would like to do all the processing on some small chip that is implanted (sending just spike events is much easier than streaming continuous data for many reasons, power and heat being two).

For my last SfN poster, we looked at various ways of filtering the continuous data to look at various bands thought to represent different classes of field potentials. In our case, it was Multi Unit Activity (MUA), which is a high frequency band, thought to encode the output of a small cortical area. We compared the ability to classify which target a person with an implanted electrode array (Utah/BrainGate 100 lead grid) was attempting to move a cursor to, which they had neural control over using unit activity. We compared the performance of the MUA signal (using two different techniques) to the traditional spike sorting, unit based way using both "poor" and "good" sorting techniques.

That's the long way of getting to this. We then also compared unit and MUA activity to what Stark and Abeles called MSU - Multiple Single Unit activity. Actually, we used two definitions of MSU. Confused yet? Let' s get past this part and it will make more sense. MSU was defined in one case as "still use the regular unit activity technique, but every waveform on an electrode belongs to the same unit. No spike sorting. Then we also tried just thresholding the continuous data, and decoding using that. So that signal is the "if the signal is over X mV, that means something is happening and therefore there is an 'event' there" technique.

Ventura's paper is a huge formalization of using the second definition of MSU, and takes it many steps further. The biggest difference being the way that she creates and separates the individual 'events', creating, in essence, surrogate 'cells'. I say 'cells' because there is no indication that they are biological, individual cells, but an activity producing lump of brain goo that fits a complicated statistical model. Actually, I don't know if it is complicated. I read through 1/2 the paper and started to glaze over (skimmed the whole thign a few times), and this is the type of paper you need some time to digest. So yes, check it out. If anything drastically different than what I said come out of it, I'll post again.

Tuesday, May 13, 2008

Sf'nN approaching

Don't forget to submit your abstracts by Thursday! I will be skipping the presentation thing this year, but I'll be there. A couple code bugs have kept me from pushing out some more MUA schtuffthz in time for the deadline (along with that annoying, looming prelim).

But, the Neurocritic has the deets on a paper in PLoS on trends in submissions to the conference. Even has pretty graphs and funky statistics.

Sunday, May 11, 2008

Uses for RSS

I am still amazed by how few people in academia use RSS. Seriously, people, get with the program. The other day I was sitting in a meeting and the PI asks, "Hey, what's (one of those labs we always watch for new papers) latest paper?" I pop up Google Reader, hit my PubMed feed for that search and start rambling off the last 5 papers. So, though I mentioned it before, here is an updated primer on RSS use in research.

First, though, why use RSS?
- You never have to navigate confusing journal sites, deal with their crappy search engines.
- You are instantly updated on all info in the field without having to do anything beyond the initial quick setup.
- All data is portable. Don't like Google Reader? Switch to another RSS reader and bring your feeds with you.
- You can build in your own way of prioritizing topics. Feeds take up no room, they don't care if you look at them every day, or once per year.
- Sharing info is remarkably easy. Sharing queries allows you to keep the whole lab on the same page. Feeds can be shared, posted on web pages, given to new lab members, created on the fly, and many other aspects that make collaborating easier.
- Archiving easily. You can mark the stories you like, and then create a feed out of those. You want the lab to read a specific paper for lab meeting or journal club each week? Just mark it. For example, I don't have everything set up perfect right now, but if you want a sneak peek at the stories I will discuss on the next Digital Trends Podcast, here's the feed.


PubMed
PubMed has the option to create custom feeds for any query. Here is how I have it set up, and highly recommend it. There are 3 phases to PubMed RSS optimization:

1) General Topic Tracking
PubMed's sloppy search engine can annoying. Say you want to keep up with motor cortex papers. Searching "motor cortex" returns anything with "motor" and "cortex", which includes molecular papers about protein expression, etc. That's okay. You could always learn a thing or two from an off-topic paper. AND remember that you are not obligated to read ANYTHING in the feed. You can tap the "Mark all as read" button and poof, they're gone.
I use several general searches that I go through last when checking my feeds. I think they're still valuable to have, since every once in a while they catch a gem and might stimulate the creatively looking at questions from multiple angles.

2) Specific Topic Tracking
Alright, the nitty gritty of PubMed. No, don't bother to learn all the syntax. Just hit the "Limits" tab, which lets you set up filters. Here is where you will spend most of your time. If you have played around with the general search, and have around 5-10 important, recent papers that you know were essential, start tweaking your search so that all of those papers show up in the results. Don't confine yourself to one search, either. You might have papers on, say, motor cortex and movement trajectory decoding, and you might have another set on LFP activity in any cortical area. Fine. You have two feeds you want to create. One "Motor Cortex Decoding" and one "Cortical LFP". Don't be afraid to make too many feeds, since you are looking to make this section as specific as possible, but also as accurate.
Yes, PubMed can suck if you don't know the details of how it searches, but you can take the time to do a very specific search with the correct wording because it will be the last time you ever use that search again.

3) Important Author/Lab Tracking
Of all the searches I do, this has been the most useful. You know the labs you want to watch. Note the most important folks (PIs, and any collaborators). Search for each set of specific authors (1 per lab), so something like "Donoghue JP[Auth] OR Hochberg LR[Auth]" (no quotes). Create a feed for each.


Individual journals
Almost all journals have the option to subscribe via RSS. You can keep up with the articles if you want, or just have the feed handy rather than having to navigate through the journal's site. "Hey, did you see that paper by Monosov in Neuron last month?" I just went from this browser window for this post to the answer in two clicks. THAT is powerful.


Individual sites
Like DNI, Digital Trends, TechRIVET, Ben's Bargains, CNet News, BoingBoing, and MetaFilter? Check them every day? Wouldn't it be easier if they all appeared in one place and you never had to hit up the site if there wasn't anything new? RSS readers will tell you what's new, let you categorize your feeds, flag stories with keywords, and make every aspect of your browsing easier.


Coordinating with others
In Google Reader at least, there is the option to share stories and attach notes about them. "Hey, check out the methods section. I think we could use that for our blahblah work." No coding up html, copying and pasting, emailing, etc.


Offline and online and via email
There are many RSS readers, both in standalone and online format. I'm not going to review them here, but I like Google Reader, since the Google Gears component allows you to take it offline as well. You can also have your feeds sent to your inbox using Feedburner or RSSFWD. Outlook 2007 has built in RSS support as well.


I went from 2-3 hours of active browsing per day for tech and research to around 10 minutes. I have returned to around 2 hours of browsing per day, but now I read around 1300 stories/day (by "read" I mean read the title, decide if I want to read the story, etc. I actually read about 20-25 full actual stories per day.) For example, I haven't hit my "General" folder yet today...
Anyhow, back to work...

Wednesday, May 7, 2008

Congrats Jodi!


And another Brown kiddo has become Dr. Kiddo. Congrats to Jodi Gilman for successfully defending her thesis from the armies of ignorance a couple weeks back.

I held off on the congrats to kill two birds with one stone. The second bird? She was interviewed regarding her research for NPR's Science Friday! Sweet. I decided to wait until the episode made it to podcast, since I missed it, before posting.

Here's the main site. Here's the story along with video. Subscribe to the NPR Science Friday Podcast here*. The topic is her paper in J Neuroscience on alcohol and reward (PubMed Linky).

*NOTE: They divide the podcast into separate files for each story. Jodi is in the 4th story - the main, longer one. I had to specifically download it in iTunes, since it is set to download the "latest episode" only, so check the drop down list.

(I'm going through an NPR phase - I looooove This American Life. I highly recommend "The Super". And, no, I haven't had a chance to see the TV show.)

Tuesday, May 6, 2008

Free wheelin'

Here are the articles related to wheelchairs that have been clogging the tubes...



First, the laser guided docking wheelchair.

Next, the NOA wheelchair. Discussed more here.


The Pegasus wheelchair concept can extend to a near standing position, which I have been told is not comfortable for people

A semi-old New Scientist article on EEG controlled wheelchairs. (Medgadget has a video.)

And lastly, extreme wheelchair sports (I feel like I posted this already, but it was still in my reader.)

The missing link


Behold, the memristor. Why is this not bigger news? Being touted as the fourth entry to holy trinity of electronics today (resistor, capacitor, inductor), the memristor retains information about its state without power.

Now stop. Seriously think about that for about 30 seconds. This means no boot up times, no spinning platters, zero power consumption sleep mode, no phantom power sucking consumer electronics, essentially on demand activity - you flip the switch and you're instantly at the previous state. Just like a light bulb. I would not be surprised to see a new power box and wall plugs for homes, where a central power distribution unit that is sent an 'on' signal over the line, that kicks electricity only to the devices that need it. More safe (juice not flowing all over the house to cause fires or zap unsuspecting lil ones), more efficient (all wire has some resistance), and it jsut plain makes sense.

Right now the EE software doesn't take advantage of this basic, elemental part, but I am excited to see what incredible devices will be built in the future. I wonder how long until the process is commercialized and available everywhere. With this discovery, HP essentially vaulted themselves into the mindspace once only occupied by IBM. (HP's had some great research, but this kicked it up a level.)

Oh, and for BCIs, the biggest problem for future real-world applications if we figured out all the crazy decoding stuff, is power. Tada! Solved.

Then again, our (USA) energy problem = solved, for now. Even if it costs more to produce, which are you going to buy: The huge honkin HDTV and PC that suck down $150/year when off, or the memristor based versions that cost $200 more?

Sorry, this is one of those news items that totally gets me juiced. Heh, juiced. I'm funny to me when I regress.

Support!

I know, no posts and then I post something utterly unrelated.

But, if you are a part of net culture, you owe it to yourself to download the new Nine Inch Nails album, which is free. No, not as in "find it on bit torrent" free, but as in on Trent's site, with a the note: "as a thank you to our fans for your continued support, we are giving away the new nine inch nails album one hundred percent free, exclusively via nin.com."

So, yes, go get it in any of several formats, including better than CD quality WAV, MP3, lossless, etc. Even if you download it and delete it, show your support for changing the music industry through REAL dedication to the art, not that namby pamby marketing stunt crap that Radiohead pulled. Reznor has been single-handedly turning the industry on its head, and this is just the latest. And yes, he is still making money doing it.

Alright, back to the BCI schuffthz....