A mindless mistake on a monotonous task may feel like a momentary glitch, but its mental roots run deep.

In a study published today in the Proceedings of the National Academy of Sciences, researchers used fMRI machines to record neurological patterns preceding careless errors.

The recordings revealed a cascade of shifting activity in the parts of the brain associated with focusing attention and maintaining routines. Researchers observed test subjects' minds going on autopilot up to half a minute before the subjects actually made mistakes, even though the subjects weren't aware of their own lapses of attention.

If the same mechanisms produce other, more meaningful errors -- slips on the assembly line or behind a steering wheel -- then the research could be used to design biofeedback systems that could catch mistakes before they're made.

"People could be made aware that they're not in the best condition to be working. Or people might learn to identify their 'bad' brain state," said study co-author Tom Eichele, a neuroscientist at the University of Bergen in Norway.

Up to 30 seconds before Eichele's test subjects carelessly said that an arrow pointing in one direction was pointing in another, blood flow decreased in their posterior medial frontal cortex, a brain region associated with sustaining effort and focus.

At the same time, activity increased in the so-called default mode network -- a region of the brain spanning the precuneus, retrosplenial cortex and anterior medial frontal cortex. The default mode network is associated with maintaining baseline routines, and tends to be most active during sleep and sedation.

In short, the conscious brain started to shut down while the system usually responsible for preventing that failed.

"This matches with our subjective perception of making mistakes on a boring task," said Michael Fox, a neuroscientist at Washington University in St. Louis who was not involved in the study. "As time goes on, you get more and more bored, and that builds up until you screw up. This study shows that scientifically."

Earlier research had identified activity in these areas prior to lapses of attention, but within just a second or two of the lapse rather than a full half-minute.

Some errors will always result from random last-second malfunction, said Eichele, but enough seem produced by the shifts he observed to raise the possibility of preventing them. Whether this can be done depends partly on whether making a mistake in reporting an arrow's direction is analogous to everyday errors, such as running a stop sign or forgetting to install a crucial bolt on an assembly line.

"Most of the things I think about in terms of decision-making in the real world have more degrees of freedom, but the same type of signals might be involved," Eichele said.

Eichele next hopes to study errors in people of different age groups, genders and personality types in a more relevant context, such as a virtual driving simulation. He also hopes to correlate the fMRI readings with electroencephalographs -- brain-based electrical patterns detectable on the scalp.

If that can be done, the brain changes that precede errors could be monitored with a few detachable electrodes rather than an expensive, automobile-sized fMRI machine. Send the readings wirelessly to a device that sets off alarms when mental attention wanders, and straying focus could be restored before it does harm.

Skin-based brain-machine interfaces are already used medically and even in videogames.

Fox suggested that, if scalp-based readings prove unreliable, researchers might look for other telltale signs of shifting attention, such as pupil dilation.

Could attention be restored through direct stimulation rather than indirect reminders? A direct-to-brain implant is highly unlikely, said Eichele.

"In principle it's possible, but I don't think it's anywhere close to being allowable," he said, citing ethical and safety issues. Even if it could prevent errors, he continued, "I don't want a wire in my brain."

---

If you like the page share it with your friends on Facebook