Catch up on stories from the past week (and beyond) at the Slashdot story archive


Forgot your password?
Medicine Science Technology

Predicting Human Errors From Brain Activity 123

Hugh Pickens writes "Researchers report that brain activity can be used to predict the likelihood of someone making an error about six seconds in advance, with gradual changes starting as much as 30 seconds ahead of time. The team used an imaging machine to scan the brains of a group of volunteers who performed a task in the presence of distracting information. When performing correctly the volunteers' brains showed increased levels of activity in those parts associated with cognitive effort, as would be expected. However, these areas gradually became less active before errors were made and at the same time another set of regions in the brain became more active. These regions are part of a so-called "default mode network" and show increased use when people are resting or asleep [PDF]. While imaging machines are far too big and complex to be used in workplaces to monitor the brain activity of people engaged in important tasks, the team hopes to correlate errors to changes in electrical activity in the brain with electroencephalography (EEG), using electrodes placed on the scalp. If EEG features can be found that correspond to the change in brain activity, then a hat that gives warning of an imminent mistake might one day become reality. We've previously discussed similar studies of brain activity."
This discussion has been archived. No new comments can be posted.

Predicting Human Errors From Brain Activity

Comments Filter:
  • Re:I agree (Score:5, Interesting)

    by GeffDE ( 712146 ) on Thursday April 24, 2008 @10:24PM (#23193300)
    Parent makes sense intuitively, but there are parts of the brain that are very sensitive to conflict; the anterior cingulate cortex (ACC) is one of these.

    If you have ever take part of a Stroop test, your ACC has been activated. In a Stroop test, the word for a color is printed in a different color i.e. the word green is shown in the color red. A participant is asked to say either the word or the color. As the speed of doing these discriminations increases, so do errors; interestingly, cognizance of errors is nearly instantaneous, however. You know that you made an error, even before the neural circuitry committed to speaking the words has finished forming the words.

    The ACC becomes more active in Stroop tests because Stroop tests cause conflicts in two neural circuits. The ACC arbitrates these circuits. Therefore, an increase in ACC activity (which will happen in advance of the error occurring) correlates with an increase in likelihood of mistakes...more in-depth research and some algorithms (I haven't RTFA) means that an error can be predicted, but of course, not with 100% success.
  • by Ungrounded Lightning ( 62228 ) on Thursday April 24, 2008 @11:05PM (#23193598) Journal
    The descriptions of "the error-detecting hat" look like the intent is to train the brain to stay alert and not make errors - or to refocus it when it wanders. (DING DING DING! HEY! WAKE UP! PAY ATTENTION!) The hat may be useful, but that use of the feedback may be the wrong approach.

    The signature they're describing corresponds, not just to a lack of alertness, but specifically a lack of alertness because the brain is going into a resting state. Seems to me that might be because all this decision-making has made the working regions of the brain tired and the brain is trying to clear them out so they'll operate properly again. So the problem is not the lack of alertness, but the attempt to continue to make decisions during the resting cycle.

    Given that, a better use of the feedback might be to tell the wearer that it's time to stop making important decisions and take a break, rather than trying to overuse a "mental muscle" that's exhausted - and perhaps train him to recognize the mental state himself so he can then dispense with the hat.

    The "break times" in working days were set up when studies showed that taking breaks, despite the "work time lost", resulted in more and better work in the work time remaining. This looks like a way to optimize the process, rather than running breaks on a clock.
  • Over-Reaching a Bit (Score:3, Interesting)

    by Nemus ( 639101 ) <> on Thursday April 24, 2008 @11:17PM (#23193674) Journal
    I have to agree with other people who've posted that this is over-reaching a little bit. This is nothing more than a correllational study: all they can say is that these systems, in this particular series of trials, showed increased activity shortly before an error took place.

    The main problem as I see it is that since they can't determine causaltiy, and only conducted this experiment with a small sample population, and with a specific task, is it could have been the task itself causing the particular regions to become active after a certain period of time. I just gave the article a quick look through, but I'd be curious to see if the errors came in distinct, set intervals. It could be simply the nature of the task that caused the activity. Furthermore, what about left handed participants? What about age groups outside of the twenties (which are a particular cohort, and can be expected to have similiar results/activity as such)? It seems like they failed to counterbalance either their participants or their trials in any meaningful way.

    Also, I'm not familiar with this journal or whatever it is, but I've never seen one where the methods section came last, which is a little strange. That's almost always the first thing I go to after the abstract.
  • Walking (Score:2, Interesting)

    by Iamthecheese ( 1264298 ) on Friday April 25, 2008 @12:00AM (#23193908)
    There are activities that can be handed off to the unconscious mind. If you are doing something so often that it seems like you could do it in your sleep, well, there's no reason not to hand it off to a subprocess and think about more important stuff.

    I think they are detecting abortive hand-off attempts: A training process for a different part of the brain than one's conscious mind. More complex motions or actions require more training. I think what we need isn't more concentration, but more error tolerance.
  • by Pooua ( 265915 ) on Friday April 25, 2008 @12:40AM (#23194132) Homepage
    I've noticed that before I make certain mistakes, or completely forget what I was doing, I will experience a spike in confusion. Sometimes, the ramp up the spike is slow, taking most of an hour; other times, it comes on me suddenly, taking only a minute. I feel confused and doing simple tasks are difficult. Then, it passes. Often, I will have completely forgotten what it was I was attempting to do, until much later when something reminds me.
  • Re:I agree (Score:3, Interesting)

    by GeffDE ( 712146 ) on Friday April 25, 2008 @07:48AM (#23196100)
    But there are some subset of errors that are known to be errors when they are performed. Your analysis is for a certain type of error that, realistically, can't be predicted. So why bother. Here's the usage scenario for this type of "error" prediction...

    You are using a computer, and you are presented with a dialog to either delete a file or cancel the dialog. You do not want to delete the file, but you click the delete button anyway. In your brain, before your finger clicks the mouse button (because the neural and muscular circuits responsible for setting that in motion take some time), you know you made an error, but you are already committed to the action and cannot stop it. A machine monitoring your brain will know you do not mean the input you are providing and will therefore disregard it, saving your file.

    If you want to be ornery about something, be ornery with the vagueness of the headline...but then again, the headline is just about as vague and sensational as all /. (and even regular news) headlines.

New systems generate new problems.