It's a cruel season that makes you get ready for bed while it's light out.

February 3, 2010 3:27 PM

Vegetated State conversations: To find out whether a simple conversation was possible, the researchers selected one of the four - a 29-year-old man who had been in a car crash. They asked him to imagine playing tennis if he wanted to answer yes to questions such as: Do you have any sisters? Is your father's name Thomas? Is your father's name Alexander? And if the answer to a question was no, he had to imagine moving round his home.
posted by bigmusic (22 comments total)

Chilling--one more for the nightmare list. I would definitely be thinking about tennis if they asked me whether I wanted physician assisted suicide.
posted by Admiral Haddock at 3:32 PM on February 3, 2010

Oh, christ.
posted by turgid dahlia at 3:34 PM on February 3, 2010

This story was on Radiolab in Sept.
posted by Lutoslawski at 3:34 PM on February 3, 2010

Do you want to go on living?

*plays tennis vigorously*

How are you planning to pay for all this care then?

*whistles past graveyard*
posted by localroger at 3:38 PM on February 3, 2010

Double-esque. (This link is much better, though.)

PS: It's vegetative.
posted by Sys Rq at 3:39 PM on February 3, 2010

I thought Dr. Steven Laureys sounded familiar, and sure enough this is the same neurologist who diagnosed Rom Houben with locked-in syndrome.

All of this fMRI evidence on locked-in syndrome seems to be coming out of the same lab or group of labs. I'd like to see some corroborating publications from groups who are not aligned with Dr. Laureys.
posted by muddgirl at 3:39 PM on February 3, 2010 [4 favorites]

More Ouija board BS?
posted by Artw at 3:44 PM on February 3, 2010

More Ouija board BS?

Well, no, because doctors and caregivers are monitoring brainwaves.
posted by KokuRyu at 3:49 PM on February 3, 2010 [1 favorite]

The new results show that you don't need behavioural indications to identify awareness and even a degree of cognitive proficiency. All you need to do is tap into brain activity directly.

Brain activity is behaviour, just very obscure behaviour.

I remember as a kid seeing an episode of Star Trek where the original captain of the enterprise was in a dalek-like wheel chair and could only say "yes" or "no" (by activating one of two lights). At the time I accepted the plot device without too much thought. Watching a repeat many years later though I thought "how stupid, if they can get him to output two variables they should be able to rig up whatever can run on binary code". It would be a tremendous amount of work to turn this into a speech synthesiser or the like (an MRI is obviously not a practical interface), as it would to learn to use it, but I'm pretty sure a lot of work has been done in related areas.
posted by GeckoDundee at 3:53 PM on February 3, 2010

So, was Terri Schaivo conscious, or not?

Wait ... did you hear that? I swear that sounded just like a bunch of long knives being sharpened...
posted by Cool Papa Bell at 4:05 PM on February 3, 2010 [1 favorite]

No need a brain to have brain activity.
posted by Artw at 4:14 PM on February 3, 2010

So say caveman neurologist.
posted by Artw at 4:15 PM on February 3, 2010 [9 favorites]

So, how reliable is this, given the "dead salmon" experiment (or whatever fish it was) that seemed to show this type of response in the brain of a dead fish?
posted by Joakim Ziegler at 4:30 PM on February 3, 2010 [1 favorite]

Not that I'm criticizing this particular use of fMRI but...

So, as the fish sat in the scanner, they showed it “a series of photographs depicting human individuals in social situations..."
posted by porpoise at 4:31 PM on February 3, 2010 [4 favorites]

So, how reliable is this, given the "dead salmon" experiment (or whatever fish it was) that seemed to show this type of response in the brain of a dead fish?

I don't know but as I'm preparing lemon-ginger salmon at the very moment, I can assure you the "tasty salmon" experiment was a smashing success.
posted by joe lisboa at 4:34 PM on February 3, 2010 [2 favorites]

It's good that the salmon experiment has popularized some of the ambiguities inherent in fMRI research, but it is easy to take it too far. In this case, the man "answered" 5 / 6 questions. Maybe that's a salmon effect. But all 5 were answered correctly. The probability of a fish randomly answering those questions correctly is 1/32, even if you assume the researchers are idiots and not properly controlling for noise. Not inconceivable, but it's evidence that they're on to something. You of course have priors too. The guy was diagnosed as vegetative, so you can't take that 1/32 straight. On the other hand, people do wake up occasionally. So it's complicated. But don't just think "dead fish". That's too easy and narrow-minded.

One should keep in mind that consciousness is not a binary proposition. It's very unlikely that this guy is a normal person trapped in an unresponsive body. Possible, but more likely his consciousness itself is highly compromised. He may still have some there though, so maybe it's worth more effort to make him comfortable, keep him company.

Finally, for many scientists, "behavior" is a loaded word. It's easy to get your ass chewed out for saying it when someone else thinks you're misusing the word. Generally, it corresponds to gross movement of an organism, although different people draw the line in different places. I personally don't give a crap as long as people are clear what they're talking about, but some people get bent out of shape on this one.
posted by Humanzee at 4:57 PM on February 3, 2010 [3 favorites]

So, how reliable is this, given the "dead salmon" experiment (or whatever fish it was) that seemed to show this type of response in the brain of a dead fish?

The lesson from the fish in the MRI is that you need to make sure your statistical model (and processing pipeline) is good at filtering out noise.

With live models, this is actually easier, since you can have some (ostensible) other state to compare against; Like resting, eyes closed vs. active, eyes open. It is worth noting however, that even in otherwise healthy research subjects, there are changes in activation during resting state even during the same study period. So, a resting state activity measurement at the beginning of the scanning process can actually be different than at the end.

So, yeah these guys may have found good evidence of locked in syndrome. But they may have also found that these people got an itchy nose or something while in the tube, too.
posted by Pogo_Fuzzybutt at 5:06 PM on February 3, 2010

Well, no, because doctors and caregivers are monitoring brainwaves.

Not brain waves, but rather activity in specific regions related to various tasks (like physical movement for tennis and special thinking for moving around the house)

So, how reliable is this, given the "dead salmon" experiment (or whatever fish it was) that seemed to show this type of response in the brain of a dead fish?

Did you actually read the FPP? The guy 'answered' 5/6 questions without the scientists having any idea what the correct answers were. The problem though is that the family got to decide if the answer was correct, rather then a third party, or sealed envelope so the family could have had their grading influenced by a desire for the guy to be conscious.
posted by delmoi at 6:02 PM on February 3, 2010

Thank you Humanzee for detailing why I'm not particularly criticizing this use of fMRI.

My general beef is with science journalism - and people who use fMRI and make extraordinary claims based on shoddy fMRI data.

Still, though, 5/6? How many times out of 20 does is it 5/6 correct? Are the family members blinded (ask Mr. Coma yes/no question, ask family member yes/no question [does it have to be a family member consensus or just a majority or just a single concurrence?] and then see it they match up)?

Personally, if it was being like this guy or being taken off life support/given "too much" painkiller - yes. Stop supplying my brain with oxygen and glucose. Or at least, ask me if that's what I want.
posted by porpoise at 9:38 PM on February 3, 2010

I agree, this would be a lot more impressive if the family had to answer the same questions blind. Also, why stop at 6? It seems like a simple and quick experiment to do, and it would be a lot more convincing if it was 19/20 or 49/50. Also, repeat it a few times. It's not like the patient has anything better to do.
posted by Joakim Ziegler at 11:39 PM on February 3, 2010

The article.

I think that the concerns about 5/6 not being meaningful (or a "multiple comparisons" issue since there were 52 patients) are lessened by them observing the brain activity specified in Part A only after asking the questions.

Also, the questions appear to have been pretty factual and verifiable "Is your father's name Alexander?" "Is your father's name Thomas?" "Do you have any brothers?" "Do you have any sisters?" are 4 of the 6 given on the figure 3. You have to read the appendix to get more of the methods; they've thought through this.
posted by a robot made out of meat at 9:41 AM on February 4, 2010

Are the family members blinded?

I think the procedure was:

(1) Blind researchers ask patient 6 questions, recording fMRI data.
(2) Second set of blind researchers examine data collected after each question (without knowing question?) and judge whether it is "tennis" or "house" (ie, yes or no).
(3) Family is asked same set of questions.
(4) A positive result is if family and second set of researchers agree. Negative result is if their results disagree.
posted by muddgirl at 9:46 AM on February 4, 2010