Remember meForgot password?
    Log in with Twitter

article imageDigital brain imaging identifies ‘suicidal thoughts’

By Tim Sandle     Nov 1, 2017 in Science
Advances with digital brain imaging can help medical professionals identify people with major depression who appear to have suicidal thoughts, enabling medical care to be delivered more speedily.
The application of the new technology is to analyze the alterations in how the brains of individuals alter in reaction to certain concepts like death, cruelty and trouble. The aim is to support existing protocols for assessing psychiatric disorders.
This is achieved through the use of digital brain scans and the utilization of machine-learning algorithms. This process provides a neural representation of specific concepts (like ‘death’, as indicated above) that are related to suicide. In other words, the research indicates that a person with suicidal tendencies can be identified if this person dwells on topics linked to death.
The technology has been developed by Carnegie Mellon University, with the research project led by Dr. Marcel Just together with Dr. David Brent from the University of Pittsburgh.
To show how effective the technology is, Bioscience Technology reports that the researchers showed lists of 10 death-related words, 10 words relating to positive concepts (such as ‘happy’) and 10 words related to negative ideas (like ‘worry’) to two groups.
In each group there were seventeen subjects, each of whom had been identified as potentially suicidal tendencies and seventeen people not showing any such tendencies (who were referred to as neurotypical individuals). When the individuals studied the words they were linked to a brain scanner.
During the study, the scientists used a machine-learning algorithm and applied this to six word-concepts that appeared to discriminate between the two groups. These words were: death, cruelty, trouble, carefree, good and praise. Using the brain representations of these six concepts, the machine learning tool could identify, at a level of 91 percent accuracy, if a participant was someone who had been identified as suicidal.
Further studies were then conducted to determine whether the algorithm was able to identify subjects who had made a suicide attempt in the past, as distinct from subjects who had only thought about it. The algorithm could do this with 94 percent accuracy.
This provides vital information to clinicians for making informed decisions. This is possible because the machine learning application interprets the brain imaging to identify concepts that lead to brain activation signatures.
The findings have been published in the journal Nature Human Behaviour. The research is titled “Machine learning of neural representations of suicide and emotion concepts identifies suicidal youth.”
More about brain imaging, Suicide, Brain scan, digital images
More news from