Everyone has probably witnessed at least one of the following: the guy at the bar who picks a fight at the slightest provocation, or the driver who explodes with rage at a tailgater. New research is beginning to more precisely locate the abnormalities in the brain that underlie this kind of violence and aggression. The findings could be used to help clinicians diagnose children and adolescents with behavioral problems, and to help clinicians tailor treatments to prevent the cycle of violence from starting. But the findings also raise thorny ethical issues: the ability to read the risk of violence in the brain could be used to stigmatize or even condemn youths before they’ve committed a crime. Alternatively, the findings could be used to make a case that criminals should not be held responsible for their behavior.

“A point will come ultimately when you could screen children and say, to a certain degree of predictability, which will become violent offenders,” says Adrian Raine, a neuroscientist at the University of Pennsylvania who studies the neurological basis of violence. “Do we do something to intervene? I think we need to start thinking about these issues now.”

In a study presented this week at the Society for Neuroscience conference in San Diego, researchers used functional magnetic brain imaging to study brain activity in a small group of adolescent boys deemed “reactively aggressive”–meaning they consistently overreact to perceived threats. “These kids tend to overreact: they punch someone or kick a door, but afterwards, they regret it,” says Guido Frank, a scientist and physician at the University of California, San Diego, who led the study. “In the moment, they can’t control themselves.”

When shown images of threatening faces, the aggressive boys had, compared with controls, greater activity in the amygdala, part of the brain that has been linked to fear, and lower activity in the prefrontal cortex, part of the brain involved in reasoning and decision making. The findings seem to provide a neurobiological explanation for their behavior: the affected adolescents feel more fearful when looking at the angry faces, as reflected in the overactive amygdala, but they may have less capacity to control their actions, due to the sluggish prefrontal cortex. “At the time, they may not be thinking about the consequences,” says Frank.

The findings build on both previous and new research implicating the prefrontal cortex in aggression and violence. In small studies of murderers and people with antisocial behavior, Raine and his colleagues found that their prefrontal cortices were smaller than those of controls. A meta-analysis–also presented at the conference–of 47 different brain-imaging studies of adults confirmed those findings: people with antisocial behavior, particularly those with a history of violent behavior, had both structural and functional impairments in that part of the brain. The prefrontal cortex was both smaller and less active in this group.

The research raises both hope and concern among scientists. Brain-imaging data can only predict risk, so it’s difficult to determine how to use it. “As we start to understand the neurobiology of violence and aggression, we have to understand that none of these factors are deterministic,” says Craig Ferris, a neuroscientist who studies aggression at Northeastern University. “We are not a slave to our biology.”

Ferris worries that searching for neurological signs of violence in children with no behavior problems could stigmatize them. “Any screening in kids is a disaster,” he says. Instead, he supports efforts to help children who already have early signs of behavioral problems. “We should use these tools to help diagnose and treat the disorders.”

It’s not yet clear how these brain abnormalities come about. Previous research has shown that genetics predominately accounts for the size of the prefrontal cortex. But abuse in infancy and childhood may also contribute. Shaken baby syndrome, for example, seems to primarily affect the orbital prefrontal cortex, one of the brain areas implicated in Raine’s study.

However, previous research in animals and humans suggests that environmental influences can have a strong impact on the ultimate outcome. Strong maternal, or other support can reduce the risk for violence in susceptible individuals, while stress and abuse can increase it. Frank hopes that his findings will ultimately aid in treating aggressive adolescents. He suggests that brain imaging might be used in conjunction with therapy to monitor an individual’s progress. “I’m a strong believer that we can change the biology and the behavior,” says Frank, who is also a psychotherapist.