Functional Magnetic Resonance Imaging (fMRI) can show thought patterns associated with cognitive processes, but can fMRI identify suicidal thoughts? Researchers at Carnegie Mellon University, where thought identification using fMRI was first developed, believe that it is certainly possible.
A recent study, in collaboration with researchers at the University of Pittsburgh, suggests that it is possible to use fMRI to identify suicidal thoughts. While the fMRI scans did not show how likely a patient is to commit suicide, thought patterns were identified that allowed the researchers to determine which patients had previously attempted suicide – and with considerable accuracy.
For the study, the researchers conducted fMRI scans on 34 patients. Half of the group were considered to be neurotypical, whereas the other half had previously been identified as having suicidal tendencies. All patients were presented with keywords which they were required to think about while undergoing the scan. Some of those keywords were intended to show the biggest difference between neurotypical patients and those with suicidal tendencies – death, trouble, good, praise, carefree and cruelty.
When patients were asked to think about each of those keywords, the researchers identified patterns in both groups of patients. The researchers were able to identify which group the patient came from with 91% accuracy based on the fMRI images taken while the patients were thinking about those six keywords. Patterns were observed for both positive keywords – such as carefree- and negative keywords such as death.
Some individuals may think about suicide, but not all individuals would attempt to take their own life. The researchers assessed whether it was possible to distinguish this and were able to predict, with 94% accuracy, which of the 17 patients had previously attempted suicide from those who had just thought about it.
While the research study does certainly suggest that it is possible to use fMRI to identify suicidal thoughts, the researchers are treating the results with caution. The study was performed on a small number of patients, and much larger studies would be required to determine how effective this brain imaging technique is at identifying suicidal patients. But as CMU’s Marcel Just explained, FMRI could be “complementary to the conventional behaviorally psychiatric diagnostic methods.”
The paper – Machine learning of neural representations of suicide and emotion concepts identifies suicidal youth – was recently published in the journal Nature Human Behavior.