Computer System Detects Faked Pain: Study

Hypochondriacs better watch out. According to new research, technology has advanced to the point where a computer system can detect when someone is faking pain better than a trained human observer.

A study published in Current Biology reveals details about a newly developed computer system that can tell the difference between spontaneous facial expressions -- such as a grimace of pain -- and voluntary ones.

According to the study, researchers compared the computer system's ability to detect faked expressions of pain with the ability of the standard person. Even after training to help them detect false facing, human subjects were only about 55 percent accurate in determine a true expression of pain verses an faked one.  The computer, however, was able to distinguish fake expressions of pain from real ones with 85 percent accuracy.

According to the researchers who designed the computer system, it's all about the mouth. Apparently people open their mouths too regularly and constantly when faking an emotion, compared to real instances of spontaneous facial reaction.

People, the researchers explain, don't normally focus on minute differences like the mouth alone, as we rely on other facial aspects such as furrowed brows and squinted eyes to normally assess the emotions of another person. After human participants were trained in ways to notice the same patterns that the computer watches for, they proved more effective than a 50-50 guess, but were still too regularly distracted by other facial aspects and frequently missed the telling cues.

According to the study's authors, the researchers hope to move on to determining how effectively detection of "over-regularity" can detected false-facing of other spontaneous emotions as well, such as grief, sleepiness, or even comprehension.

The study was published in Current Biology on March 20.